Re: svn commit: r359374 - in head: . share/man/man4 share/man/man7 share/man/man9 sys/crypto/aesni sys/crypto/armv8 sys/crypto/blake2 sys/crypto/ccp sys/crypto/via sys/dev/cesa sys/dev/cxgbe sys/dev/c
On 4/13/20 12:31 PM, Warner Losh wrote: > On Mon, Apr 13, 2020 at 1:22 PM John-Mark Gurney wrote: > >> John Baldwin wrote this message on Mon, Apr 13, 2020 at 09:56 -0700: >>> On 4/12/20 1:08 PM, Xin Li wrote: >>>> >>>> >>>> On 3/27/20 11:25 AM, John Baldwin wrote: >>>> [...]> - Drivers no longer register a list of supported algorithms. >> This >>>>> doesn't quite work when you factor in modes (e.g. a driver might >>>>> support both AES-CBC and SHA2-256-HMAC separately but not combined >>>>> for ETA). Instead, a new 'crypto_probesession' method has been >>>>> added to the kobj interface for symmteric crypto drivers. This >>>>> method returns a negative value on success (similar to how >>>>> device_probe works) and the crypto framework uses this value to >> pick >>>>> the "best" driver. There are three constants for hardware >>>>> (e.g. ccr), accelerated software (e.g. aesni), and plain software >>>>> (cryptosoft) that give preference in that order. One effect of >> this >>>>> is that if you request only hardware when creating a new session, >>>>> you will no longer get a session using accelerated software. >>>>> Another effect is that the default setting to disallow software >>>>> crypto via /dev/crypto now disables accelerated software. >>>> >>>> For user-visible interface, it seems like we are essentially treating >>>> "accelerated software" like AES-NI the same way of plain software. For >>>> example, geom_eli would now say: >>>> >>>> GEOM_ELI: Encryption: AES-XTS 128 >>>> GEOM_ELI: Crypto: software >>>> >>>> Instead of: >>>> >>>> GEOM_ELI: Encryption: AES-XTS 128 >>>> GEOM_ELI: Crypto: hardware >>>> >>>> When AES-NI is used (which is because we only have two bits to >> represent >>>> hardware and software, and have gave neither bits clear with its own >>>> meaning (use specific driver)). >>>> >>>> If we are not going to add a new bit to represent accelerated software, >>>> why are they categorized as software providers? Technically, all these >>>> still requires hardware that implements the cryptographic primitives to >>>> work, and it's much easier for system administrators if we expose the >>>> fact that they are using some kind of acceleration than asking them to >>>> run DTrace etc. to find out. Personally, I think it's probably better >>>> to change the notion to either "accelerated" (by either hardware or >>>> software) and "software"... >>> >>> The only case where this is visible is in fact GELI (there is no printf >>> for IPsec or KTLS). For /dev/crypto using aesni.ko is a bug, not a >>> feature, as any such software would be much better off using AES-NI in >>> userland instead of round-tripping through the kernel. We could add a >>> bit to appease the GELI printf, or we could just kill the GELI >>> printf. I think a more useful approach would probably be to kill the >>> GELI printf and instead add something less GELI-specific such as >>> counters of active sessions for the various cryptographic drivers that >>> would show which drivers are in use for any in-kernel crypto. This >>> approach also lends itself to supporting a more flexible API where a >>> single crypto session might be backed by multiple drivers where a >>> binary hardware / software setting might not even make sense as you >>> might have a mix. (I know of other out-of-tree crypto use cases that >>> experimented with splitting in-kernel crypto workloads between an >>> async co-processor and AES-NI.) >> >> As long as there remains some what to warn the user that when they've >> mounted a geli volume that they're using slow crypto, I'm fine... >> >> We have done a bad job on doing the right thing, and I'm afraid that >> removing this print w/o doing something to address it will go from >> FreeBSD already not doing the right thing, but to not allowing the user >> to know that FreeBSD isn't doing the right thing. >> >> Even a simple print in the crypto driver when the processor supports >> AES-NI, but the aesni module isn't loaded would be useful.. >> >> Without the geli print, it's likely articles, like the recent A
Re: svn commit: r359374 - in head: . share/man/man4 share/man/man7 share/man/man9 sys/crypto/aesni sys/crypto/armv8 sys/crypto/blake2 sys/crypto/ccp sys/crypto/via sys/dev/cesa sys/dev/cxgbe sys/dev/c
On 4/12/20 1:08 PM, Xin Li wrote: > > > On 3/27/20 11:25 AM, John Baldwin wrote: > [...]> - Drivers no longer register a list of supported algorithms. This >> doesn't quite work when you factor in modes (e.g. a driver might >> support both AES-CBC and SHA2-256-HMAC separately but not combined >> for ETA). Instead, a new 'crypto_probesession' method has been >> added to the kobj interface for symmteric crypto drivers. This >> method returns a negative value on success (similar to how >> device_probe works) and the crypto framework uses this value to pick >> the "best" driver. There are three constants for hardware >> (e.g. ccr), accelerated software (e.g. aesni), and plain software >> (cryptosoft) that give preference in that order. One effect of this >> is that if you request only hardware when creating a new session, >> you will no longer get a session using accelerated software. >> Another effect is that the default setting to disallow software >> crypto via /dev/crypto now disables accelerated software. > > For user-visible interface, it seems like we are essentially treating > "accelerated software" like AES-NI the same way of plain software. For > example, geom_eli would now say: > > GEOM_ELI: Encryption: AES-XTS 128 > GEOM_ELI: Crypto: software > > Instead of: > > GEOM_ELI: Encryption: AES-XTS 128 > GEOM_ELI: Crypto: hardware > > When AES-NI is used (which is because we only have two bits to represent > hardware and software, and have gave neither bits clear with its own > meaning (use specific driver)). > > If we are not going to add a new bit to represent accelerated software, > why are they categorized as software providers? Technically, all these > still requires hardware that implements the cryptographic primitives to > work, and it's much easier for system administrators if we expose the > fact that they are using some kind of acceleration than asking them to > run DTrace etc. to find out. Personally, I think it's probably better > to change the notion to either "accelerated" (by either hardware or > software) and "software"... The only case where this is visible is in fact GELI (there is no printf for IPsec or KTLS). For /dev/crypto using aesni.ko is a bug, not a feature, as any such software would be much better off using AES-NI in userland instead of round-tripping through the kernel. We could add a bit to appease the GELI printf, or we could just kill the GELI printf. I think a more useful approach would probably be to kill the GELI printf and instead add something less GELI-specific such as counters of active sessions for the various cryptographic drivers that would show which drivers are in use for any in-kernel crypto. This approach also lends itself to supporting a more flexible API where a single crypto session might be backed by multiple drivers where a binary hardware / software setting might not even make sense as you might have a mix. (I know of other out-of-tree crypto use cases that experimented with splitting in-kernel crypto workloads between an async co-processor and AES-NI.) Also, while AES-NI instructions are faster than plain C, they are still very much software rather than an engine like QAT or ccr(4). They run synchronously using host CPU cycles rather than async on a separate co-processor. This means that if you tweak the various geli sysctls for batching vs non-batching, etc. you need to make the same choices for both accelerated software and "plain" software vs possibly different choices for an async co-processor. -- John Baldwin ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r359788 - head
Author: jhb Date: Fri Apr 10 23:13:29 2020 New Revision: 359788 URL: https://svnweb.freebsd.org/changeset/base/359788 Log: Document removal of deprecated algorithms for in-kernel GSS. Modified: head/RELNOTES Modified: head/RELNOTES == --- head/RELNOTES Fri Apr 10 23:10:28 2020(r359787) +++ head/RELNOTES Fri Apr 10 23:13:29 2020(r359788) @@ -10,6 +10,10 @@ newline. Entries should be separated by a newline. Changes to this file should not be MFCed. +r359786-r359787: + Remove support for DES, Triple DES, and RC4 from in-kernel GSS + authentication. + r357627: remove elf2aout. ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r359787 - head/usr.sbin/gssd
Author: jhb
Date: Fri Apr 10 23:10:28 2020
New Revision: 359787
URL: https://svnweb.freebsd.org/changeset/base/359787
Log:
Remove the -o option from gssd(8).
This uses DES and the kernel no longer supports DES for in-kernel GSS.
Reviewed by: kp
Relnotes: yes
Sponsored by: Chelsio Communications
Differential Revision:https://reviews.freebsd.org/D24345
Modified:
head/usr.sbin/gssd/gssd.8
head/usr.sbin/gssd/gssd.c
Modified: head/usr.sbin/gssd/gssd.8
==
--- head/usr.sbin/gssd/gssd.8 Fri Apr 10 23:08:41 2020(r359786)
+++ head/usr.sbin/gssd/gssd.8 Fri Apr 10 23:10:28 2020(r359787)
@@ -25,7 +25,7 @@
.\"
.\" $FreeBSD$
.\"
-.Dd July 7, 2013
+.Dd April 10 2020
.Dt GSSD 8
.Os
.Sh NAME
@@ -35,7 +35,6 @@
.Nm
.Op Fl d
.Op Fl h
-.Op Fl o
.Op Fl v
.Op Fl s Ar dir-list
.Op Fl c Ar file-substring
@@ -61,10 +60,6 @@ gssname
option for the
.Xr mount_nfs 8
command.
-.It Fl o
-Force use of DES and the associated old style GSS-API initialization token.
-This may be required to make kerberized NFS mounts work against some
-non-FreeBSD NFS servers.
.It Fl v
Run in verbose mode.
In this mode,
Modified: head/usr.sbin/gssd/gssd.c
==
--- head/usr.sbin/gssd/gssd.c Fri Apr 10 23:08:41 2020(r359786)
+++ head/usr.sbin/gssd/gssd.c Fri Apr 10 23:10:28 2020(r359787)
@@ -77,7 +77,6 @@ int debug_level;
static char ccfile_dirlist[PATH_MAX + 1], ccfile_substring[NAME_MAX + 1];
static char pref_realm[1024];
static int verbose;
-static int use_old_des;
static int hostbased_initiator_cred;
#ifndef WITHOUT_KERBEROS
/* 1.2.752.43.13.14 */
@@ -125,7 +124,7 @@ main(int argc, char **argv)
pref_realm[0] = '\0';
debug = 0;
verbose = 0;
- while ((ch = getopt(argc, argv, "dhovs:c:r:")) != -1) {
+ while ((ch = getopt(argc, argv, "dhvs:c:r:")) != -1) {
switch (ch) {
case 'd':
debug_level++;
@@ -142,17 +141,6 @@ main(int argc, char **argv)
" without MK_KERBEROS\n");
#endif
break;
- case 'o':
-#ifndef WITHOUT_KERBEROS
- /*
-* Force use of DES and the old type of GSSAPI token.
-*/
- use_old_des = 1;
-#else
- errx(1, "This option not available when built"
- " without MK_KERBEROS\n");
-#endif
- break;
case 'v':
verbose = 1;
break;
@@ -483,44 +471,6 @@ gssd_init_sec_context_1_svc(init_sec_context_args *arg
}
gotcred = 0;
-#ifndef WITHOUT_KERBEROS
- if (use_old_des != 0) {
- if (cred == GSS_C_NO_CREDENTIAL) {
- /* Acquire a credential for the uid. */
- maj_stat = gssd_get_user_cred(_stat, argp->uid,
- );
- if (maj_stat == GSS_S_COMPLETE)
- gotcred = 1;
- else
- gssd_verbose_out("gssd_init_sec_context: "
- "get user cred failed uid=%d major=0x%x "
- "minor=%d\n", (int)argp->uid,
- (unsigned int)maj_stat, (int)min_stat);
- }
- if (cred != GSS_C_NO_CREDENTIAL) {
- key_enctype = ETYPE_DES_CBC_CRC;
- enctype[0] = (key_enctype >> 24) & 0xff;
- enctype[1] = (key_enctype >> 16) & 0xff;
- enctype[2] = (key_enctype >> 8) & 0xff;
- enctype[3] = key_enctype & 0xff;
- principal_desc.length = sizeof(enctype);
- principal_desc.value = enctype;
- result->major_status = gss_set_cred_option(
- >minor_status, ,
- GSS_KRB5_SET_ALLOWABLE_ENCTYPES_X,
- _desc);
- gssd_verbose_out("gssd_init_sec_context: set allowable "
- "enctype major=0x%x minor=%d\n",
- (unsigned int)result->major_status,
- (int)result->minor_status);
- if (result->major_status != GSS_S_COMPLETE) {
- if (gotcred != 0)
- gss_release_cred(_stat, );
- return (TRUE);
- }
- }
- }
-#endif
result->major_status = gss_init_sec_context(>minor_status,
cred, , name, argp->mech_type,
argp->req_flags, argp->time_req,
svn commit: r359786 - in head/sys: conf kgssapi/krb5 modules/kgssapi_krb5
Author: jhb
Date: Fri Apr 10 23:08:41 2020
New Revision: 359786
URL: https://svnweb.freebsd.org/changeset/base/359786
Log:
Remove support for Kernel GSS algorithms deprecated in r348875.
This removes support for using DES, Triple DES, and RC4.
Reviewed by: cem, kp
Tested by:kp
Sponsored by: Chelsio Communications
Differential Revision:https://reviews.freebsd.org/D24344
Deleted:
head/sys/kgssapi/krb5/kcrypto_arcfour.c
head/sys/kgssapi/krb5/kcrypto_des.c
head/sys/kgssapi/krb5/kcrypto_des3.c
Modified:
head/sys/conf/files
head/sys/kgssapi/krb5/kcrypto.c
head/sys/kgssapi/krb5/kcrypto.h
head/sys/modules/kgssapi_krb5/Makefile
Modified: head/sys/conf/files
==
--- head/sys/conf/files Fri Apr 10 22:42:14 2020(r359785)
+++ head/sys/conf/files Fri Apr 10 23:08:41 2020(r359786)
@@ -3945,9 +3945,6 @@ kgssapi/gssd_prot.c optional kgssapi
kgssapi/krb5/krb5_mech.c optional kgssapi
kgssapi/krb5/kcrypto.c optional kgssapi
kgssapi/krb5/kcrypto_aes.c optional kgssapi
-kgssapi/krb5/kcrypto_arcfour.c optional kgssapi
-kgssapi/krb5/kcrypto_des.c optional kgssapi
-kgssapi/krb5/kcrypto_des3.coptional kgssapi
kgssapi/kgss_if.m optional kgssapi
kgssapi/gsstest.c optional kgssapi_debug
# These files in libkern/ are those needed by all architectures. Some
Modified: head/sys/kgssapi/krb5/kcrypto.c
==
--- head/sys/kgssapi/krb5/kcrypto.c Fri Apr 10 22:42:14 2020
(r359785)
+++ head/sys/kgssapi/krb5/kcrypto.c Fri Apr 10 23:08:41 2020
(r359786)
@@ -42,19 +42,10 @@ __FBSDID("$FreeBSD$");
#include "kcrypto.h"
static struct krb5_encryption_class *krb5_encryption_classes[] = {
- _des_encryption_class,
- _des3_encryption_class,
_aes128_encryption_class,
_aes256_encryption_class,
- _arcfour_encryption_class,
- _arcfour_56_encryption_class,
NULL
};
-
-struct timeval krb5_warn_interval = { .tv_sec = 3600, .tv_usec = 0 };
-SYSCTL_TIMEVAL_SEC(_kern, OID_AUTO, kgssapi_warn_interval, CTLFLAG_RW,
-_warn_interval,
-"Delay in seconds between warnings of deprecated KGSSAPI crypto.");
struct krb5_encryption_class *
krb5_find_encryption_class(int etype)
Modified: head/sys/kgssapi/krb5/kcrypto.h
==
--- head/sys/kgssapi/krb5/kcrypto.h Fri Apr 10 22:42:14 2020
(r359785)
+++ head/sys/kgssapi/krb5/kcrypto.h Fri Apr 10 23:08:41 2020
(r359786)
@@ -95,13 +95,8 @@ struct krb5_key_state {
void*ks_priv;
};
-extern struct krb5_encryption_class krb5_des_encryption_class;
-extern struct krb5_encryption_class krb5_des3_encryption_class;
extern struct krb5_encryption_class krb5_aes128_encryption_class;
extern struct krb5_encryption_class krb5_aes256_encryption_class;
-extern struct krb5_encryption_class krb5_arcfour_encryption_class;
-extern struct krb5_encryption_class krb5_arcfour_56_encryption_class;
-extern struct timeval krb5_warn_interval;
static __inline void
krb5_set_key(struct krb5_key_state *ks, const void *keydata)
Modified: head/sys/modules/kgssapi_krb5/Makefile
==
--- head/sys/modules/kgssapi_krb5/Makefile Fri Apr 10 22:42:14 2020
(r359785)
+++ head/sys/modules/kgssapi_krb5/Makefile Fri Apr 10 23:08:41 2020
(r359786)
@@ -5,10 +5,7 @@ KMOD= kgssapi_krb5
SRCS= krb5_mech.c \
kcrypto.c \
- kcrypto_des.c \
- kcrypto_des3.c \
kcrypto_aes.c \
- kcrypto_arcfour.c \
opt_inet6.h
SRCS+= kgss_if.h gssd.h
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r359784 - in head/sys/dev/cxgbe: . crypto
On 4/10/20 3:27 PM, John Baldwin wrote: > Author: jhb > Date: Fri Apr 10 22:27:45 2020 > New Revision: 359784 > URL: https://svnweb.freebsd.org/changeset/base/359784 > > Log: > Use both crypto engines on a T6. > > A T6 adapter contains two crypto engines on separate channels. This > commit distributes sessions between the two engines. Previously, only > the first engine was used. In some simple benchmarks using KTLS with the traffic flowing over one of the ports (cc0) of the same NIC ccr0 was on, throughput increased from ~41 Gbps to 45 Gbps with this change (albeit with CPU usage increase from 70% to 80%). In addition, this commit adds a 'dev.ccr..port_mask' sysctl which controls which ports are enabled for new sessions. Setting this to '2' during this test to force all crypto requests to use the crypto engine associated with the channel for cc1 (which was not being used to send any traffic) further increased the throughput to around 48 Gbps while using the same amount of CPU. -- John Baldwin ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r359784 - in head/sys/dev/cxgbe: . crypto
Author: jhb
Date: Fri Apr 10 22:27:45 2020
New Revision: 359784
URL: https://svnweb.freebsd.org/changeset/base/359784
Log:
Use both crypto engines on a T6.
A T6 adapter contains two crypto engines on separate channels. This
commit distributes sessions between the two engines. Previously, only
the first engine was used.
Reviewed by: np
Sponsored by: Chelsio Communications
Differential Revision:https://reviews.freebsd.org/D24347
Modified:
head/sys/dev/cxgbe/adapter.h
head/sys/dev/cxgbe/crypto/t4_crypto.c
Modified: head/sys/dev/cxgbe/adapter.h
==
--- head/sys/dev/cxgbe/adapter.hFri Apr 10 22:18:30 2020
(r359783)
+++ head/sys/dev/cxgbe/adapter.hFri Apr 10 22:27:45 2020
(r359784)
@@ -600,9 +600,7 @@ struct sge_rxq {
struct sge_fl fl; /* MUST follow iq */
struct ifnet *ifp; /* the interface this rxq belongs to */
-#if defined(INET) || defined(INET6)
struct lro_ctrl lro;/* LRO state */
-#endif
/* stats for common events first */
Modified: head/sys/dev/cxgbe/crypto/t4_crypto.c
==
--- head/sys/dev/cxgbe/crypto/t4_crypto.c Fri Apr 10 22:18:30 2020
(r359783)
+++ head/sys/dev/cxgbe/crypto/t4_crypto.c Fri Apr 10 22:27:45 2020
(r359784)
@@ -162,10 +162,18 @@ struct ccr_session_blkcipher {
char deckey[CHCR_AES_MAX_KEY_LEN];
};
+struct ccr_port {
+ struct sge_wrq *txq;
+ struct sge_rxq *rxq;
+ int tx_channel_id;
+ u_int active_sessions;
+};
+
struct ccr_session {
bool active;
int pending;
enum { HASH, HMAC, BLKCIPHER, ETA, GCM, CCM } mode;
+ struct ccr_port *port;
union {
struct ccr_session_hmac hmac;
struct ccr_session_gmac gmac;
@@ -178,11 +186,10 @@ struct ccr_softc {
struct adapter *adapter;
device_t dev;
uint32_t cid;
- int tx_channel_id;
struct mtx lock;
bool detaching;
- struct sge_wrq *txq;
- struct sge_rxq *rxq;
+ struct ccr_port ports[MAX_NPORTS];
+ u_int port_mask;
/*
* Pre-allocate S/G lists used when preparing a work request.
@@ -297,7 +304,8 @@ ccr_phys_dsgl_len(int nsegs)
}
static void
-ccr_write_phys_dsgl(struct ccr_softc *sc, void *dst, int nsegs)
+ccr_write_phys_dsgl(struct ccr_softc *sc, struct ccr_session *s, void *dst,
+int nsegs)
{
struct sglist *sg;
struct cpl_rx_phys_dsgl *cpl;
@@ -316,7 +324,7 @@ ccr_write_phys_dsgl(struct ccr_softc *sc, void *dst, i
V_CPL_RX_PHYS_DSGL_PCITPHNTENB(0) | V_CPL_RX_PHYS_DSGL_DCAID(0) |
V_CPL_RX_PHYS_DSGL_NOOFSGENTR(nsegs));
cpl->rss_hdr_int.opcode = CPL_RX_PHYS_ADDR;
- cpl->rss_hdr_int.qid = htobe16(sc->rxq->iq.abs_id);
+ cpl->rss_hdr_int.qid = htobe16(s->port->rxq->iq.abs_id);
cpl->rss_hdr_int.hash_val = 0;
sgl = (struct phys_sge_pairs *)(cpl + 1);
j = 0;
@@ -392,9 +400,9 @@ ccr_use_imm_data(u_int transhdr_len, u_int input_len)
}
static void
-ccr_populate_wreq(struct ccr_softc *sc, struct chcr_wr *crwr, u_int kctx_len,
-u_int wr_len, u_int imm_len, u_int sgl_len, u_int hash_size,
-struct cryptop *crp)
+ccr_populate_wreq(struct ccr_softc *sc, struct ccr_session *s,
+struct chcr_wr *crwr, u_int kctx_len, u_int wr_len, u_int imm_len,
+u_int sgl_len, u_int hash_size, struct cryptop *crp)
{
u_int cctx_size, idata_len;
@@ -409,13 +417,13 @@ ccr_populate_wreq(struct ccr_softc *sc, struct chcr_wr
V_FW_CRYPTO_LOOKASIDE_WR_LEN16(wr_len / 16));
crwr->wreq.session_id = 0;
crwr->wreq.rx_chid_to_rx_q_id = htobe32(
- V_FW_CRYPTO_LOOKASIDE_WR_RX_CHID(sc->tx_channel_id) |
+ V_FW_CRYPTO_LOOKASIDE_WR_RX_CHID(s->port->tx_channel_id) |
V_FW_CRYPTO_LOOKASIDE_WR_LCB(0) |
V_FW_CRYPTO_LOOKASIDE_WR_PHASH(0) |
V_FW_CRYPTO_LOOKASIDE_WR_IV(IV_NOP) |
V_FW_CRYPTO_LOOKASIDE_WR_FQIDX(0) |
V_FW_CRYPTO_LOOKASIDE_WR_TX_CH(0) |
- V_FW_CRYPTO_LOOKASIDE_WR_RX_Q_ID(sc->rxq->iq.abs_id));
+ V_FW_CRYPTO_LOOKASIDE_WR_RX_Q_ID(s->port->rxq->iq.abs_id));
crwr->wreq.key_addr = 0;
crwr->wreq.pld_size_hash_size = htobe32(
V_FW_CRYPTO_LOOKASIDE_WR_PLD_SIZE(sgl_len) |
@@ -424,8 +432,9 @@ ccr_populate_wreq(struct ccr_softc *sc, struct chcr_wr
crwr->ulptx.cmd_dest = htobe32(V_ULPTX_CMD(ULP_TX_PKT) |
V_ULP_TXPKT_DATAMODIFY(0) |
- V_ULP_TXPKT_CHANNELID(sc->tx_channel_id) | V_ULP_TXPKT_DEST(0) |
- V_ULP_TXPKT_FID(sc->rxq->iq.abs_id) | V_ULP_TXPKT_RO(1));
+ V_ULP_TXPKT_CHANNELID(s->port->tx_channel_id) |
+ V_ULP_TXPKT_DEST(0) |
+ V_ULP_TXPKT_FID(s->port->rxq->iq.abs_id) |
svn commit: r359570 - head/sys/opencrypto
Author: jhb
Date: Thu Apr 2 17:22:16 2020
New Revision: 359570
URL: https://svnweb.freebsd.org/changeset/base/359570
Log:
Avoid checking pointers that are never NULL.
Coverity noted that cod pointer is always non-NULL at the end of
cryptodev_aead(). While here, fix cryptodev_op() to match by making
one earlier failure case before cod and crp are allocated just return
directly.
CID: 1422185
Reported by: Coverity
Modified:
head/sys/opencrypto/cryptodev.c
Modified: head/sys/opencrypto/cryptodev.c
==
--- head/sys/opencrypto/cryptodev.c Thu Apr 2 16:55:28 2020
(r359569)
+++ head/sys/opencrypto/cryptodev.c Thu Apr 2 17:22:16 2020
(r359570)
@@ -915,8 +915,7 @@ cryptodev_op(
if (cop->mac && cse->hashsize == 0) {
SDT_PROBE1(opencrypto, dev, ioctl, error, __LINE__);
- error = EINVAL;
- goto bail;
+ return (EINVAL);
}
/*
@@ -1091,10 +1090,8 @@ again:
}
bail:
- if (crp)
- crypto_freereq(crp);
- if (cod)
- cod_free(cod);
+ crypto_freereq(crp);
+ cod_free(cod);
return (error);
}
@@ -1285,8 +1282,7 @@ again:
bail:
crypto_freereq(crp);
- if (cod)
- cod_free(cod);
+ cod_free(cod);
return (error);
}
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r359569 - head/sys/opencrypto
Author: jhb Date: Thu Apr 2 16:55:28 2020 New Revision: 359569 URL: https://svnweb.freebsd.org/changeset/base/359569 Log: Remove duplicate conditional. CID: 1422192 Reported by: Coverity Modified: head/sys/opencrypto/crypto.c Modified: head/sys/opencrypto/crypto.c == --- head/sys/opencrypto/crypto.cThu Apr 2 16:48:46 2020 (r359568) +++ head/sys/opencrypto/crypto.cThu Apr 2 16:55:28 2020 (r359569) @@ -722,8 +722,6 @@ alg_is_digest(int alg) return (true); if (alg >= CRYPTO_SHA2_256_HMAC && alg <= CRYPTO_SHA2_512_HMAC) return (true); - if (alg >= CRYPTO_SHA2_256_HMAC && alg <= CRYPTO_SHA2_512_HMAC) - return (true); if (alg == CRYPTO_AES_NIST_GMAC) return (true); if (alg >= CRYPTO_BLAKE2B && alg <= CRYPTO_BLAKE2S) ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r359524 - head/sys/kgssapi/krb5
Author: jhb Date: Wed Apr 1 17:09:21 2020 New Revision: 359524 URL: https://svnweb.freebsd.org/changeset/base/359524 Log: Set crp_ilen for crypto requests. Assertions in crypto_dispatch() depend on this value being set to verify that payload and AAD regions are in bounds. Also, requests that use a single kernel buffer rely on this to know how long the buffer is for bus_dma, etc. Reported by: kp Modified: head/sys/kgssapi/krb5/kcrypto_aes.c Modified: head/sys/kgssapi/krb5/kcrypto_aes.c == --- head/sys/kgssapi/krb5/kcrypto_aes.c Wed Apr 1 17:08:55 2020 (r359523) +++ head/sys/kgssapi/krb5/kcrypto_aes.c Wed Apr 1 17:09:21 2020 (r359524) @@ -158,6 +158,7 @@ aes_encrypt_1(const struct krb5_key_state *ks, int buf crp->crp_buf_type = buftype; crp->crp_buf = buf; + crp->crp_ilen = skip + len; crp->crp_opaque = as; crp->crp_callback = aes_crypto_cb; @@ -329,6 +330,7 @@ aes_checksum(const struct krb5_key_state *ks, int usag crp->crp_flags = CRYPTO_F_CBIFSYNC; crp->crp_buf_type = CRYPTO_BUF_MBUF; crp->crp_mbuf = inout; + crp->crp_ilen = skip + inlen + 12; crp->crp_opaque = as; crp->crp_callback = aes_crypto_cb; ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r359530 - in head: . lib/libsysdecode sys/amd64/amd64 sys/amd64/ia32 sys/arm/arm sys/arm64/arm64 sys/compat/ia32 sys/conf sys/fs/procfs sys/i386/i386 sys/kern sys/mips/mips sys/modules/...
Author: jhb
Date: Wed Apr 1 19:22:09 2020
New Revision: 359530
URL: https://svnweb.freebsd.org/changeset/base/359530
Log:
Retire procfs-based process debugging.
Modern debuggers and process tracers use ptrace() rather than procfs
for debugging. ptrace() has a supserset of functionality available
via procfs and new debugging features are only added to ptrace().
While the two debugging services share some fields in struct proc,
they each use dedicated fields and separate code. This results in
extra complexity to support a feature that hasn't been enabled in the
default install for several years.
PR: 244939 (exp-run)
Reviewed by: kib, mjg (earlier version)
Relnotes: yes
Differential Revision:https://reviews.freebsd.org/D23837
Deleted:
head/sys/fs/procfs/procfs_ioctl.c
head/sys/sys/pioctl.h
Modified:
head/ObsoleteFiles.inc
head/lib/libsysdecode/Makefile
head/sys/amd64/amd64/trap.c
head/sys/amd64/amd64/vm_machdep.c
head/sys/amd64/ia32/ia32_reg.c
head/sys/amd64/ia32/ia32_signal.c
head/sys/amd64/ia32/ia32_syscall.c
head/sys/arm/arm/syscall.c
head/sys/arm64/arm64/trap.c
head/sys/compat/ia32/ia32_sysvec.c
head/sys/conf/files
head/sys/fs/procfs/procfs.c
head/sys/fs/procfs/procfs.h
head/sys/i386/i386/trap.c
head/sys/i386/i386/vm_machdep.c
head/sys/kern/imgact_elf.c
head/sys/kern/kern_exec.c
head/sys/kern/kern_exit.c
head/sys/kern/kern_fork.c
head/sys/kern/kern_prot.c
head/sys/kern/kern_sig.c
head/sys/kern/kern_thread.c
head/sys/kern/subr_syscall.c
head/sys/kern/subr_trap.c
head/sys/kern/sys_process.c
head/sys/mips/mips/trap.c
head/sys/modules/procfs/Makefile
head/sys/powerpc/powerpc/trap.c
head/sys/riscv/riscv/trap.c
head/sys/sys/param.h
head/sys/sys/proc.h
Modified: head/ObsoleteFiles.inc
==
--- head/ObsoleteFiles.inc Wed Apr 1 19:09:47 2020(r359529)
+++ head/ObsoleteFiles.inc Wed Apr 1 19:22:09 2020(r359530)
@@ -36,6 +36,9 @@
# xargs -n1 | sort | uniq -d;
# done
+# 20200401: Remove procfs-based process debugging
+OLD_FILES+=usr/include/sys/pioctl.h
+
# 20200330: GDB_LIBEXEC option retired (always true)
OLD_FILES+=usr/bin/gdb
OLD_FILES+=usr/bin/gdbserver
Modified: head/lib/libsysdecode/Makefile
==
--- head/lib/libsysdecode/Makefile Wed Apr 1 19:09:47 2020
(r359529)
+++ head/lib/libsysdecode/Makefile Wed Apr 1 19:22:09 2020
(r359530)
@@ -119,11 +119,6 @@ CFLAGS+=-DPF
# Workaround duplicate declarations in
CFLAGS.gcc.ioctl.c+= -Wno-redundant-decls
-# Ignore deprecation warning in
-CFLAGS.clang.ioctl.c+= -Wno-\#warnings
-CFLAGS.gcc.ioctl.c+= -Wno-cpp
-
-CFLAGS.clang+= ${CFLAGS.clang.${.IMPSRC}}
CFLAGS.gcc+= ${CFLAGS.gcc.${.IMPSRC}}
DEPENDOBJS+= tables.h
Modified: head/sys/amd64/amd64/trap.c
==
--- head/sys/amd64/amd64/trap.c Wed Apr 1 19:09:47 2020(r359529)
+++ head/sys/amd64/amd64/trap.c Wed Apr 1 19:22:09 2020(r359530)
@@ -57,7 +57,6 @@ __FBSDID("$FreeBSD$");
#include
#include
#include
-#include
#include
#include
#include
Modified: head/sys/amd64/amd64/vm_machdep.c
==
--- head/sys/amd64/amd64/vm_machdep.c Wed Apr 1 19:09:47 2020
(r359529)
+++ head/sys/amd64/amd64/vm_machdep.c Wed Apr 1 19:22:09 2020
(r359530)
@@ -58,7 +58,6 @@ __FBSDID("$FreeBSD$");
#include
#include
#include
-#include
#include
#include
#include
@@ -198,15 +197,11 @@ cpu_fork(struct thread *td1, struct proc *p2, struct t
td2->td_frame->tf_rdx = 1;
/*
-* If the parent process has the trap bit set (i.e. a debugger had
-* single stepped the process to the system call), we need to clear
-* the trap flag from the new frame unless the debugger had set PF_FORK
-* on the parent. Otherwise, the child will receive a (likely
-* unexpected) SIGTRAP when it executes the first instruction after
-* returning to userland.
+* If the parent process has the trap bit set (i.e. a debugger
+* had single stepped the process to the system call), we need
+* to clear the trap flag from the new frame.
*/
- if ((p1->p_pfsflags & PF_FORK) == 0)
- td2->td_frame->tf_rflags &= ~PSL_T;
+ td2->td_frame->tf_rflags &= ~PSL_T;
/*
* Set registers for trampoline to user mode. Leave space for the
Modified: head/sys/amd64/ia32/ia32_reg.c
==
--- head/sys/amd64/ia32/ia32_reg.c Wed Apr 1 19:09:47 2020
(r359529)
+++ head/sys/amd64/ia32/ia32_reg.c Wed Apr 1 19:22:09 2020
(r359530)
Re: svn commit: r359459 - head/share/man/man5
On 3/30/20 1:08 PM, Ed Maste wrote: > Author: emaste > Date: Mon Mar 30 20:08:26 2020 > New Revision: 359459 > URL: https://svnweb.freebsd.org/changeset/base/359459 > > Log: > regen src.conf.5 after WITHOUT_GDB_LIBEXEC removal r359454 > > ... and a number of other changes since the last regen. Hmm, 359454 was a markup fix to elf.5 and I haven't seen the commit to actually remove GDB_LIBEXEC yet? -- John Baldwin ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r359467 - head/lib/libc/sys
Author: jhb Date: Mon Mar 30 21:48:47 2020 New Revision: 359467 URL: https://svnweb.freebsd.org/changeset/base/359467 Log: Trim some duplicate EIO descriptions. While here, drop an extra conjunction from the list of error conditions for the remaining EIO description in symlink(2). Discussed with: mckusick (trimming duplicates) MFC after:2 weeks Modified: head/lib/libc/sys/mkfifo.2 head/lib/libc/sys/symlink.2 Modified: head/lib/libc/sys/mkfifo.2 == --- head/lib/libc/sys/mkfifo.2 Mon Mar 30 21:44:30 2020(r359466) +++ head/lib/libc/sys/mkfifo.2 Mon Mar 30 21:48:47 2020(r359467) @@ -130,10 +130,6 @@ which the fifo is being created has been exhausted. An .Tn I/O error occurred while making the directory entry or allocating the inode. -.It Bq Er EIO -An -.Tn I/O -error occurred while reading from or writing to the file system. .It Bq Er EINTEGRITY Corrupted data was detected while reading from the file system. .It Bq Er EFAULT Modified: head/lib/libc/sys/symlink.2 == --- head/lib/libc/sys/symlink.2 Mon Mar 30 21:44:30 2020(r359466) +++ head/lib/libc/sys/symlink.2 Mon Mar 30 21:48:47 2020(r359467) @@ -118,7 +118,7 @@ manual page for more information. .It Bq Er EIO An I/O error occurred while making the directory entry for .Fa name2 , -or allocating the inode for +allocating the inode for .Fa name2 , or writing out the link contents of .Fa name2 . @@ -149,8 +149,6 @@ contain the symbolic link has been exhausted. .It Bq Er EDQUOT The user's quota of inodes on the file system on which the symbolic link is being created has been exhausted. -.It Bq Er EIO -An I/O error occurred while making the directory entry or allocating the inode. .It Bq Er EINTEGRITY Corrupted data was detected while reading from the file system. .It Bq Er EFAULT ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r359465 - in head: lib/libc/sys share/man/man9
Author: jhb Date: Mon Mar 30 21:44:00 2020 New Revision: 359465 URL: https://svnweb.freebsd.org/changeset/base/359465 Log: Document EINTEGRITY errors for many system calls. EINTEGRITY was previously documented as a UFS-specific error for mount(2). This documents EINTEGRITY as a filesystem-independent error that may be reported by the backing store of a filesystem. While here, document EIO as a filesystem-independent error for both mount(2) and posix_fadvise(2). EIO was previously only documented for UFS for mount(2). Reviewed by: mckusick Suggested by: mckusick MFC after:2 weeks Sponsored by: Chelsio Communications Differential Revision:https://reviews.freebsd.org/D24168 Modified: head/lib/libc/sys/access.2 head/lib/libc/sys/acct.2 head/lib/libc/sys/bind.2 head/lib/libc/sys/chdir.2 head/lib/libc/sys/chflags.2 head/lib/libc/sys/chmod.2 head/lib/libc/sys/chown.2 head/lib/libc/sys/chroot.2 head/lib/libc/sys/copy_file_range.2 head/lib/libc/sys/execve.2 head/lib/libc/sys/fhlink.2 head/lib/libc/sys/fhreadlink.2 head/lib/libc/sys/fsync.2 head/lib/libc/sys/getdirentries.2 head/lib/libc/sys/getfh.2 head/lib/libc/sys/getfsstat.2 head/lib/libc/sys/ktrace.2 head/lib/libc/sys/link.2 head/lib/libc/sys/mkdir.2 head/lib/libc/sys/mkfifo.2 head/lib/libc/sys/mknod.2 head/lib/libc/sys/mount.2 head/lib/libc/sys/open.2 head/lib/libc/sys/pathconf.2 head/lib/libc/sys/posix_fadvise.2 head/lib/libc/sys/posix_fallocate.2 head/lib/libc/sys/quotactl.2 head/lib/libc/sys/read.2 head/lib/libc/sys/readlink.2 head/lib/libc/sys/rename.2 head/lib/libc/sys/rmdir.2 head/lib/libc/sys/sendfile.2 head/lib/libc/sys/stat.2 head/lib/libc/sys/statfs.2 head/lib/libc/sys/swapon.2 head/lib/libc/sys/symlink.2 head/lib/libc/sys/truncate.2 head/lib/libc/sys/undelete.2 head/lib/libc/sys/unlink.2 head/lib/libc/sys/utimensat.2 head/lib/libc/sys/utimes.2 head/lib/libc/sys/write.2 head/share/man/man9/VOP_COPY_FILE_RANGE.9 head/share/man/man9/VOP_READDIR.9 head/share/man/man9/VOP_READLINK.9 head/share/man/man9/g_data.9 Modified: head/lib/libc/sys/access.2 == --- head/lib/libc/sys/access.2 Mon Mar 30 21:42:46 2020(r359464) +++ head/lib/libc/sys/access.2 Mon Mar 30 21:44:00 2020(r359465) @@ -28,7 +28,7 @@ .\" @(#)access.2 8.2 (Berkeley) 4/1/94 .\" $FreeBSD$ .\" -.Dd November 11, 2018 +.Dd March 30, 2020 .Dt ACCESS 2 .Os .Sh NAME @@ -175,6 +175,8 @@ argument points outside the process's allocated address space. .It Bq Er EIO An I/O error occurred while reading from or writing to the file system. +.It Bq Er EINTEGRITY +Corrupted data was detected while reading from the file system. .El .Pp Also, the Modified: head/lib/libc/sys/acct.2 == --- head/lib/libc/sys/acct.2Mon Mar 30 21:42:46 2020(r359464) +++ head/lib/libc/sys/acct.2Mon Mar 30 21:44:00 2020(r359465) @@ -28,7 +28,7 @@ .\" @(#)acct.2 8.1 (Berkeley) 6/4/93 .\" $FreeBSD$ .\" -.Dd April 17, 2004 +.Dd March 30, 2020 .Dt ACCT 2 .Os .Sh NAME @@ -119,6 +119,8 @@ argument points outside the process's allocated address space. .It Bq Er EIO An I/O error occurred while reading from or writing to the file system. +.It Bq Er EINTEGRITY +Corrupted data was detected while reading from the file system. .El .Sh SEE ALSO .Xr acct 5 , Modified: head/lib/libc/sys/bind.2 == --- head/lib/libc/sys/bind.2Mon Mar 30 21:42:46 2020(r359464) +++ head/lib/libc/sys/bind.2Mon Mar 30 21:44:00 2020(r359465) @@ -28,7 +28,7 @@ .\" @(#)bind.2 8.1 (Berkeley) 6/4/93 .\" $FreeBSD$ .\" -.Dd August 18, 2016 +.Dd March 30, 2020 .Dt BIND 2 .Os .Sh NAME @@ -122,6 +122,8 @@ A prefix component of the path name does not exist. Too many symbolic links were encountered in translating the pathname. .It Bq Er EIO An I/O error occurred while making the directory entry or allocating the inode. +.It Bq Er EINTEGRITY +Corrupted data was detected while reading from the file system. .It Bq Er EROFS The name would reside on a read-only file system. .It Bq Er EISDIR Modified: head/lib/libc/sys/chdir.2 == --- head/lib/libc/sys/chdir.2 Mon Mar 30 21:42:46 2020(r359464) +++ head/lib/libc/sys/chdir.2 Mon Mar 30 21:44:00 2020(r359465) @@ -28,7 +28,7 @@ .\" @(#)chdir.28.2 (Berkeley) 12/11/93 .\" $FreeBSD$ .\" -.Dd December 1, 2017 +.Dd March 30, 2020 .Dt CHDIR 2 .Os .Sh NAME @@ -96,6 +96,8 @@ argument points outside the process's allocated address space. .It Bq Er EIO An I/O error occurred while reading from or writing to the file system. +.It Bq Er EINTEGRITY +Corrupted
Re: svn commit: r359374 - in head: . share/man/man4 share/man/man7 share/man/man9 sys/crypto/aesni sys/crypto/armv8 sys/crypto/blake2 sys/crypto/ccp sys/crypto/via sys/dev/cesa sys/dev/cxgbe sys/dev/c
On 3/27/20 11:25 AM, John Baldwin wrote: > Author: jhb > Date: Fri Mar 27 18:25:23 2020 > New Revision: 359374 > URL: https://svnweb.freebsd.org/changeset/base/359374 > > Log: > Refactor driver and consumer interfaces for OCF (in-kernel crypto). I have tested as much as I have accessible. Someone did test cesa(4) I believe and found it worked with cryptocheck but not IPsec. I found a bug in the hash handling related to IPsec that I fixed in cesa(4), but the original submitter didn't retest. kp@ has an out-of-tree driver that he said works with these changes as well. -- John Baldwin ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r359374 - in head: . share/man/man4 share/man/man7 share/man/man9 sys/crypto/aesni sys/crypto/armv8 sys/crypto/blake2 sys/crypto/ccp sys/crypto/via sys/dev/cesa sys/dev/cxgbe sys/dev/cx...
Author: jhb Date: Fri Mar 27 18:25:23 2020 New Revision: 359374 URL: https://svnweb.freebsd.org/changeset/base/359374 Log: Refactor driver and consumer interfaces for OCF (in-kernel crypto). - The linked list of cryptoini structures used in session initialization is replaced with a new flat structure: struct crypto_session_params. This session includes a new mode to define how the other fields should be interpreted. Available modes include: - COMPRESS (for compression/decompression) - CIPHER (for simply encryption/decryption) - DIGEST (computing and verifying digests) - AEAD (combined auth and encryption such as AES-GCM and AES-CCM) - ETA (combined auth and encryption using encrypt-then-authenticate) Additional modes could be added in the future (e.g. if we wanted to support TLS MtE for AES-CBC in the kernel we could add a new mode for that. TLS modes might also affect how AAD is interpreted, etc.) The flat structure also includes the key lengths and algorithms as before. However, code doesn't have to walk the linked list and switch on the algorithm to determine which key is the auth key vs encryption key. The 'csp_auth_*' fields are always used for auth keys and settings and 'csp_cipher_*' for cipher. (Compression algorithms are stored in csp_cipher_alg.) - Drivers no longer register a list of supported algorithms. This doesn't quite work when you factor in modes (e.g. a driver might support both AES-CBC and SHA2-256-HMAC separately but not combined for ETA). Instead, a new 'crypto_probesession' method has been added to the kobj interface for symmteric crypto drivers. This method returns a negative value on success (similar to how device_probe works) and the crypto framework uses this value to pick the "best" driver. There are three constants for hardware (e.g. ccr), accelerated software (e.g. aesni), and plain software (cryptosoft) that give preference in that order. One effect of this is that if you request only hardware when creating a new session, you will no longer get a session using accelerated software. Another effect is that the default setting to disallow software crypto via /dev/crypto now disables accelerated software. Once a driver is chosen, 'crypto_newsession' is invoked as before. - Crypto operations are now solely described by the flat 'cryptop' structure. The linked list of descriptors has been removed. A separate enum has been added to describe the type of data buffer in use instead of using CRYPTO_F_* flags to make it easier to add more types in the future if needed (e.g. wired userspace buffers for zero-copy). It will also make it easier to re-introduce separate input and output buffers (in-kernel TLS would benefit from this). Try to make the flags related to IV handling less insane: - CRYPTO_F_IV_SEPARATE means that the IV is stored in the 'crp_iv' member of the operation structure. If this flag is not set, the IV is stored in the data buffer at the 'crp_iv_start' offset. - CRYPTO_F_IV_GENERATE means that a random IV should be generated and stored into the data buffer. This cannot be used with CRYPTO_F_IV_SEPARATE. If a consumer wants to deal with explicit vs implicit IVs, etc. it can always generate the IV however it needs and store partial IVs in the buffer and the full IV/nonce in crp_iv and set CRYPTO_F_IV_SEPARATE. The layout of the buffer is now described via fields in cryptop. crp_aad_start and crp_aad_length define the boundaries of any AAD. Previously with GCM and CCM you defined an auth crd with this range, but for ETA your auth crd had to span both the AAD and plaintext (and they had to be adjacent). crp_payload_start and crp_payload_length define the boundaries of the plaintext/ciphertext. Modes that only do a single operation (COMPRESS, CIPHER, DIGEST) should only use this region and leave the AAD region empty. If a digest is present (or should be generated), it's starting location is marked by crp_digest_start. Instead of using the CRD_F_ENCRYPT flag to determine the direction of the operation, cryptop now includes an 'op' field defining the operation to perform. For digests I've added a new VERIFY digest mode which assumes a digest is present in the input and fails the request with EBADMSG if it doesn't match the internally-computed digest. GCM and CCM already assumed this, and the new AEAD mode requires this for decryption. The new ETA mode now also requires this for decryption, so IPsec and GELI no longer do their own authentication verification. Simple DIGEST operations can also do this, though there are no in-tree consumers. To eventually support some refcounting to close races, the session
Re: svn commit: r359349 - head/gnu/usr.bin/binutils/ld
On 3/27/20 9:54 AM, Brandon Bergren wrote: > > > On Fri, Mar 27, 2020, at 11:40 AM, John Baldwin wrote: >> On 3/26/20 6:12 PM, Brandon Bergren wrote: >>> Author: bdragon >>> Date: Fri Mar 27 01:12:11 2020 >>> New Revision: 359349 >>> URL: https://svnweb.freebsd.org/changeset/base/359349 >>> >>> Log: >>> Now that there are no remaining users of ld.bfd, remove the ld subdir. >>> >>> Deleted: >>> head/gnu/usr.bin/binutils/ld/ >> >> I think Ed mentioned some other cleanups to do as part of this change in >> ObsoleteFiles.inc as well? > > The problem is that they're still conditional on whether or not base/binutils > is installed, so they *have* to live in OptionalObsoleteFiles.inc as far as > I'm aware. Nope. We moved the GCC ones to ObsoleteFiles.inc and kept the conditionals for the ports version in that file. -- John Baldwin ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r359349 - head/gnu/usr.bin/binutils/ld
On 3/26/20 6:12 PM, Brandon Bergren wrote: > Author: bdragon > Date: Fri Mar 27 01:12:11 2020 > New Revision: 359349 > URL: https://svnweb.freebsd.org/changeset/base/359349 > > Log: > Now that there are no remaining users of ld.bfd, remove the ld subdir. > > Deleted: > head/gnu/usr.bin/binutils/ld/ I think Ed mentioned some other cleanups to do as part of this change in ObsoleteFiles.inc as well? -- John Baldwin ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r359275 - head
Author: jhb
Date: Tue Mar 24 18:16:02 2020
New Revision: 359275
URL: https://svnweb.freebsd.org/changeset/base/359275
Log:
Disable rarely used architecture variants in make universe by default.
If EXTRA_TARGETS is defined, build all supported architecture
variants. By default, build architecture variants needed to provide
code coverage or that are commonly used.
Use this to disable building of all the hard-float and little-endian
MIPS architecture variants along with n32 by default.
Reviewed by: rpokala
Discussed with: imp, emaste
Differential Revision:https://reviews.freebsd.org/D24178
Modified:
head/Makefile
Modified: head/Makefile
==
--- head/Makefile Tue Mar 24 17:54:34 2020(r359274)
+++ head/Makefile Tue Mar 24 18:16:02 2020(r359275)
@@ -483,16 +483,17 @@ worlds: .PHONY
.if make(universe) || make(universe_kernels) || make(tinderbox) || \
make(targets) || make(universe-toolchain)
#
-# Always build architectures supported by clang. Only build architectures
-# only supported by GCC if a suitable toolchain is present or enabled.
-# In all cases, if the user specifies TARGETS on the command line,
-# honor that most of all.
+# Don't build rarely used architectures unless requested.
#
+.if defined(EXTRA_TARGETS)
+EXTRA_ARCHES_mips= mipsel mipshf mipselhf mips64el mips64hf mips64elhf
+EXTRA_ARCHES_mips+=mipsn32
+.endif
TARGETS?=amd64 arm arm64 i386 mips powerpc riscv
_UNIVERSE_TARGETS= ${TARGETS}
TARGET_ARCHES_arm?=armv6 armv7
TARGET_ARCHES_arm64?= aarch64
-TARGET_ARCHES_mips?= mipsel mips mips64el mips64 mipsn32 mipselhf mipshf
mips64elhf mips64hf
+TARGET_ARCHES_mips?= mips mips64 ${EXTRA_ARCHES_mips}
# powerpcspe excluded until clang fixed
TARGET_ARCHES_powerpc?=powerpc powerpc64
TARGET_ARCHES_riscv?= riscv64 riscv64sf
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r359263 - head/sys/geom/eli
Author: jhb
Date: Mon Mar 23 21:26:32 2020
New Revision: 359263
URL: https://svnweb.freebsd.org/changeset/base/359263
Log:
Use the newer EINTEGRITY error when authentication fails.
GELI used to fail with EINVAL when a read request spanned a disk
sector whose contents did not match the sector's authentication tag.
The recently-added EINTEGRITY more closely matches to the error in
this case.
Reviewed by: cem, mckusick
MFC after:2 weeks
Sponsored by: Chelsio Communications
Differential Revision:https://reviews.freebsd.org/D24131
Modified:
head/sys/geom/eli/g_eli_integrity.c
Modified: head/sys/geom/eli/g_eli_integrity.c
==
--- head/sys/geom/eli/g_eli_integrity.c Mon Mar 23 21:21:38 2020
(r359262)
+++ head/sys/geom/eli/g_eli_integrity.c Mon Mar 23 21:26:32 2020
(r359263)
@@ -232,7 +232,7 @@ g_eli_auth_read_done(struct cryptop *crp)
bp->bio_driver2 = NULL;
if (bp->bio_error != 0) {
if (bp->bio_error == -1)
- bp->bio_error = EINVAL;
+ bp->bio_error = EINTEGRITY;
else {
G_ELI_LOGREQ(0, bp,
"Crypto READ request failed (error=%d).",
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r357614 - in head/sys: kern sys
On 2/6/20 4:45 AM, Pawel Biernacki wrote: > Author: kaktus > Date: Thu Feb 6 12:45:58 2020 > New Revision: 357614 > URL: https://svnweb.freebsd.org/changeset/base/357614 > > Log: > sysctl(9): add CTLFLAG_NEEDGIANT flag > > Add CTLFLAG_NEEDGIANT flag (modelled after D_NEEDGIANT) that will be used to > mark sysctls that still require locking Giant. > > Rewrite sysctl_handle_string() to use internal locking instead of locking > Giant. This broke CTLFLAG_RDTUN strings such as hw.cxgbe.config_file. This is supposed to be writable by setting the associated environment variable via loader.conf or kenv: >From sys/dev/cxgbe/t4_main.c: /* * Configuration file. All the _CF names here are special. */ #define DEFAULT_CF "default" #define BUILTIN_CF "built-in" #define FLASH_CF"flash" #define UWIRE_CF"uwire" #define FPGA_CF "fpga" static char t4_cfg_file[32] = DEFAULT_CF; SYSCTL_STRING(_hw_cxgbe, OID_AUTO, config_file, CTLFLAG_RDTUN, t4_cfg_file, sizeof(t4_cfg_file), "Firmware configuration file"); However, CTLFLAG_RDTUN does not include CTLFLAG_WR, so when the kernel attempts to "write" to this node (the "TUN" part) to apply the associated kenv variable when loading the kernel module, your changes here now treat it as a ro_string and set 'arg2' to the length of the actual string. In my case I was setting the value to a longer string that still fit in the allocated space, and the value now fails to set giving me the following error on the console: Setting sysctl hw.cxgbe.config_file failed: 22 You can reproduce by doing the following: # kenv hw.cxgbe.config_file="kern_tls" # kldload if_cxgbe That should give you the error, and then checking the hw.cxgbe.config_file sysctl afterwards will show "default" instead of "kern_tls". Note that you don't need any cxgbe hardware to reproduce this as a fix would result in the value of the sysctl after the kldload being "kern_tls". -- John Baldwin ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r359053 - head/sys/kern
On 3/17/20 3:36 PM, Ian Lepore wrote:
> On Tue, 2020-03-17 at 22:27 +, Conrad Meyer wrote:
>> Author: cem
>> Date: Tue Mar 17 22:27:16 2020
>> New Revision: 359053
>> URL: https://svnweb.freebsd.org/changeset/base/359053
>>
>> Log:
>> Implement sysctl kern.boot_id
>>
>> Boot IDs are random, opaque 128-bit identifiers that distinguish distinct
>> system boots. A new ID is generated each time the system boots. Unlike
>> kern.boottime, the value is not modified by NTP adjustments. It remains
>> fixed
>> until the machine is restarted.
>>
>> PR:244867
>> Reported by: Ricardo Fraile
>> MFC after: I do not intend to, but feel free
>>
>> Modified:
>> head/sys/kern/kern_mib.c
>>
>> Modified: head/sys/kern/kern_mib.c
>> ==
>> --- head/sys/kern/kern_mib.c Tue Mar 17 21:29:03 2020(r359052)
>> +++ head/sys/kern/kern_mib.c Tue Mar 17 22:27:16 2020(r359053)
>> @@ -448,6 +448,32 @@ SYSCTL_PROC(_kern, KERN_HOSTID, hostid,
>> CTLTYPE_ULONG | CTLFLAG_RW | CTLFLAG_PRISON | CTLFLAG_MPSAFE |
>> CTLFLAG_CAPRD,
>> NULL, 0, sysctl_hostid, "LU", "Host ID");
>>
>> +static struct mtx bootid_lk;
>> +MTX_SYSINIT(bootid_lock, _lk, "bootid generator lock", MTX_DEF);
>> +
>> +static int
>> +sysctl_bootid(SYSCTL_HANDLER_ARGS)
>> +{
>> +static uint8_t boot_id[16];
>> +static bool initialized = false;
>> +
>> +mtx_lock(_lk);
>> +if (!initialized) {
>> +if (!is_random_seeded()) {
>> +mtx_unlock(_lk);
>> +return (ENXIO);
>> +}
>> +arc4random_buf(boot_id, sizeof(boot_id));
>> +initialized = true;
>> +}
>> +mtx_unlock(_lk);
>> +
>> +return (SYSCTL_OUT(req, boot_id, sizeof(boot_id)));
>> +}
>> +SYSCTL_PROC(_kern, OID_AUTO, boot_id,
>> +CTLTYPE_STRUCT | CTLFLAG_RD | CTLFLAG_MPSAFE | CTLFLAG_CAPRD,
>> +NULL, 0, sysctl_bootid, "", "Random boot ID");
>> +
>> /*
>> * The osrelease string is copied from the global (osrelease in vers.c) into
>> * prison0 by a sysinit and is inherited by child jails if not changed at
>> jail
>
> This seems a bit complex. Why run a sysinit to init a mutex so that
> you can safely do a lazy init of boot_id? Seems like it would be much
> easier to just use a sysinit at SI_SUB_LAST to init boot_id before
> sysctl can reference it.
Presumably you may not have enough entropy by SI_SUB_LAST to generate it?
--
John Baldwin
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r359054 - head/lib/libsysdecode
Author: jhb
Date: Tue Mar 17 23:22:07 2020
New Revision: 359054
URL: https://svnweb.freebsd.org/changeset/base/359054
Log:
Fix the workaround to ignore the #warning for GCC.
clang and gcc use different warning flags for #warning preprocessor
directives.
Reported by: Jenkins
MFC after:1 week
Modified:
head/lib/libsysdecode/Makefile
Modified: head/lib/libsysdecode/Makefile
==
--- head/lib/libsysdecode/Makefile Tue Mar 17 22:27:16 2020
(r359053)
+++ head/lib/libsysdecode/Makefile Tue Mar 17 23:22:07 2020
(r359054)
@@ -119,10 +119,12 @@ CFLAGS+=-DPF
# Workaround duplicate declarations in
CFLAGS.gcc.ioctl.c+= -Wno-redundant-decls
-CFLAGS.gcc+= ${CFLAGS.gcc.${.IMPSRC}}
-
# Ignore deprecation warning in
-CFLAGS.ioctl.c+= -Wno-\#warnings
+CFLAGS.clang.ioctl.c+= -Wno-\#warnings
+CFLAGS.gcc.ioctl.c+= -Wno-cpp
+
+CFLAGS.clang+= ${CFLAGS.clang.${.IMPSRC}}
+CFLAGS.gcc+= ${CFLAGS.gcc.${.IMPSRC}}
DEPENDOBJS+= tables.h
tables.h: mktables
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r359050 - head/tools/build/mk
Author: jhb
Date: Tue Mar 17 21:11:38 2020
New Revision: 359050
URL: https://svnweb.freebsd.org/changeset/base/359050
Log:
Add missing DTrace files for WITHOUT_CDDL=yes.
Reviewed by: dteske
MFC after:1 week
Sponsored by: DARPA
Differential Revision:https://reviews.freebsd.org/D24037
Modified:
head/tools/build/mk/OptionalObsoleteFiles.inc
Modified: head/tools/build/mk/OptionalObsoleteFiles.inc
==
--- head/tools/build/mk/OptionalObsoleteFiles.inc Tue Mar 17 19:53:03
2020(r359049)
+++ head/tools/build/mk/OptionalObsoleteFiles.inc Tue Mar 17 21:11:38
2020(r359050)
@@ -1151,13 +1151,18 @@ OLD_FILES+=usr/lib/dtrace/drti.o
OLD_FILES+=usr/lib/dtrace/errno.d
OLD_FILES+=usr/lib/dtrace/io.d
OLD_FILES+=usr/lib/dtrace/ip.d
+OLD_FILES+=usr/lib/dtrace/mbuf.d
OLD_FILES+=usr/lib/dtrace/psinfo.d
.if ${TARGET_ARCH} == "amd64" || ${TARGET_ARCH} == "i386"
OLD_FILES+=usr/lib/dtrace/regs_x86.d
.endif
+OLD_FILES+=usr/lib/dtrace/sctp.d
+OLD_FILES+=usr/lib/dtrace/siftr.d
OLD_FILES+=usr/lib/dtrace/signal.d
+OLD_FILES+=usr/lib/dtrace/socket.d
OLD_FILES+=usr/lib/dtrace/tcp.d
OLD_FILES+=usr/lib/dtrace/udp.d
+OLD_FILES+=usr/lib/dtrace/udplite.d
OLD_FILES+=usr/lib/dtrace/unistd.d
OLD_FILES+=usr/lib/libavl.a
OLD_FILES+=usr/lib/libavl.so
@@ -1205,29 +1210,120 @@ OLD_LIBS+=usr/lib32/libuutil.so.2
OLD_FILES+=usr/lib32/libuutil_p.a
.endif
OLD_LIBS+=lib/libdtrace.so.2
+OLD_FILES+=usr/libexec/dwatch/chmod
+OLD_FILES+=usr/libexec/dwatch/errno
+OLD_FILES+=usr/libexec/dwatch/fchmodat
+OLD_FILES+=usr/libexec/dwatch/io
+OLD_FILES+=usr/libexec/dwatch/io-done
+OLD_FILES+=usr/libexec/dwatch/io-start
+OLD_FILES+=usr/libexec/dwatch/ip
+OLD_FILES+=usr/libexec/dwatch/ip-receive
+OLD_FILES+=usr/libexec/dwatch/ip-send
+OLD_FILES+=usr/libexec/dwatch/kill
+OLD_FILES+=usr/libexec/dwatch/lchmod
+OLD_FILES+=usr/libexec/dwatch/nanosleep
+OLD_FILES+=usr/libexec/dwatch/open
+OLD_FILES+=usr/libexec/dwatch/openat
+OLD_FILES+=usr/libexec/dwatch/proc
+OLD_FILES+=usr/libexec/dwatch/proc-create
+OLD_FILES+=usr/libexec/dwatch/proc-exec
+OLD_FILES+=usr/libexec/dwatch/proc-exec-failure
+OLD_FILES+=usr/libexec/dwatch/proc-exec-success
+OLD_FILES+=usr/libexec/dwatch/proc-exit
+OLD_FILES+=usr/libexec/dwatch/proc-signal
+OLD_FILES+=usr/libexec/dwatch/proc-signal-clear
+OLD_FILES+=usr/libexec/dwatch/proc-signal-discard
+OLD_FILES+=usr/libexec/dwatch/proc-signal-send
+OLD_FILES+=usr/libexec/dwatch/proc-status
+OLD_FILES+=usr/libexec/dwatch/read
+OLD_FILES+=usr/libexec/dwatch/recv
+OLD_FILES+=usr/libexec/dwatch/recvfrom
+OLD_FILES+=usr/libexec/dwatch/recvmsg
+OLD_FILES+=usr/libexec/dwatch/rw
+OLD_FILES+=usr/libexec/dwatch/sched
+OLD_FILES+=usr/libexec/dwatch/sched-change-pri
+OLD_FILES+=usr/libexec/dwatch/sched-cpu
+OLD_FILES+=usr/libexec/dwatch/sched-dequeue
+OLD_FILES+=usr/libexec/dwatch/sched-enqueue
+OLD_FILES+=usr/libexec/dwatch/sched-exec
+OLD_FILES+=usr/libexec/dwatch/sched-lend-pri
+OLD_FILES+=usr/libexec/dwatch/sched-load-change
+OLD_FILES+=usr/libexec/dwatch/sched-off-cpu
+OLD_FILES+=usr/libexec/dwatch/sched-on-cpu
+OLD_FILES+=usr/libexec/dwatch/sched-preempt
+OLD_FILES+=usr/libexec/dwatch/sched-pri
+OLD_FILES+=usr/libexec/dwatch/sched-queue
+OLD_FILES+=usr/libexec/dwatch/sched-remain-cpu
+OLD_FILES+=usr/libexec/dwatch/sched-sleep
+OLD_FILES+=usr/libexec/dwatch/sched-surrender
+OLD_FILES+=usr/libexec/dwatch/sched-tick
+OLD_FILES+=usr/libexec/dwatch/sched-wakeup
+OLD_FILES+=usr/libexec/dwatch/send
+OLD_FILES+=usr/libexec/dwatch/sendmsg
+OLD_FILES+=usr/libexec/dwatch/sendrecv
+OLD_FILES+=usr/libexec/dwatch/sendto
+OLD_FILES+=usr/libexec/dwatch/systop
+OLD_FILES+=usr/libexec/dwatch/tcp
+OLD_FILES+=usr/libexec/dwatch/tcp-accept
+OLD_FILES+=usr/libexec/dwatch/tcp-accept-established
+OLD_FILES+=usr/libexec/dwatch/tcp-accept-refused
+OLD_FILES+=usr/libexec/dwatch/tcp-connect
+OLD_FILES+=usr/libexec/dwatch/tcp-connect-established
+OLD_FILES+=usr/libexec/dwatch/tcp-connect-refused
+OLD_FILES+=usr/libexec/dwatch/tcp-connect-request
+OLD_FILES+=usr/libexec/dwatch/tcp-established
+OLD_FILES+=usr/libexec/dwatch/tcp-init
+OLD_FILES+=usr/libexec/dwatch/tcp-io
+OLD_FILES+=usr/libexec/dwatch/tcp-receive
+OLD_FILES+=usr/libexec/dwatch/tcp-refused
+OLD_FILES+=usr/libexec/dwatch/tcp-send
+OLD_FILES+=usr/libexec/dwatch/tcp-state-change
+OLD_FILES+=usr/libexec/dwatch/tcp-status
+OLD_FILES+=usr/libexec/dwatch/udp
+OLD_FILES+=usr/libexec/dwatch/udp-receive
+OLD_FILES+=usr/libexec/dwatch/udp-send
+OLD_FILES+=usr/libexec/dwatch/udplite
+OLD_FILES+=usr/libexec/dwatch/udplite-receive
+OLD_FILES+=usr/libexec/dwatch/udplite-send
+OLD_FILES+=usr/libexec/dwatch/vop_create
+OLD_FILES+=usr/libexec/dwatch/vop_lookup
+OLD_FILES+=usr/libexec/dwatch/vop_mkdir
+OLD_FILES+=usr/libexec/dwatch/vop_mknod
+OLD_FILES+=usr/libexec/dwatch/vop_readdir
+OLD_FILES+=usr/libexec/dwatch/vop_remove
+OLD_FILES+=usr/libexec/dwatch/vop_rename
Re: svn commit: r359047 - in head: lib/libsysdecode sys/fs/procfs sys/sys
On 3/17/20 11:44 AM, John Baldwin wrote: > Author: jhb > Date: Tue Mar 17 18:44:03 2020 > New Revision: 359047 > URL: https://svnweb.freebsd.org/changeset/base/359047 > > Log: > Mark procfs-based process debugging as deprecated for FreeBSD 13. > > Attempting to use ioctls on /proc//mem to control a process will > trigger warnings on the console. The include file will > also now emit a compile-time warning when used from userland. > > Reviewed by:emaste > MFC after: 1 week > Relnotes: yes > Differential Revision: https://reviews.freebsd.org/D23822 Missed: PR: 244383 (exp-run) I also didn't quite say why. My original reason for why is that I have some debugging changes in flight to try to fix spurious EINTR's that break the GDB test suite (and we also have an open bug about spurious errors for gcore and truss). mjg@ wants to do this to remove overhead from the syscall path. Both of these reasons stem from the fact that procfs debugging is not just a layer on top of ptrace but sits beside it requiring duplicate checks in code, etc. They reuse some, but not all, of the same infrastructure. At this point, ptrace() now provides a superset of functionality relative to procfs, and that is only going to continue going forward as only ptrace() based debugging has seen active work in the past several years. -- John Baldwin ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r359047 - in head: lib/libsysdecode sys/fs/procfs sys/sys
Author: jhb
Date: Tue Mar 17 18:44:03 2020
New Revision: 359047
URL: https://svnweb.freebsd.org/changeset/base/359047
Log:
Mark procfs-based process debugging as deprecated for FreeBSD 13.
Attempting to use ioctls on /proc//mem to control a process will
trigger warnings on the console. The include file will
also now emit a compile-time warning when used from userland.
Reviewed by: emaste
MFC after:1 week
Relnotes: yes
Differential Revision:https://reviews.freebsd.org/D23822
Modified:
head/lib/libsysdecode/Makefile
head/sys/fs/procfs/procfs_ioctl.c
head/sys/sys/pioctl.h
Modified: head/lib/libsysdecode/Makefile
==
--- head/lib/libsysdecode/Makefile Tue Mar 17 18:17:32 2020
(r359046)
+++ head/lib/libsysdecode/Makefile Tue Mar 17 18:44:03 2020
(r359047)
@@ -121,6 +121,9 @@ CFLAGS.gcc.ioctl.c+= -Wno-redundant-decls
CFLAGS.gcc+= ${CFLAGS.gcc.${.IMPSRC}}
+# Ignore deprecation warning in
+CFLAGS.ioctl.c+= -Wno-\#warnings
+
DEPENDOBJS+= tables.h
tables.h: mktables
sh ${.CURDIR}/mktables ${SYSROOT:U${DESTDIR}}${INCLUDEDIR} ${.TARGET}
Modified: head/sys/fs/procfs/procfs_ioctl.c
==
--- head/sys/fs/procfs/procfs_ioctl.c Tue Mar 17 18:17:32 2020
(r359046)
+++ head/sys/fs/procfs/procfs_ioctl.c Tue Mar 17 18:44:03 2020
(r359047)
@@ -69,10 +69,53 @@ procfs_ioctl(PFS_IOCTL_ARGS)
#ifdef COMPAT_FREEBSD6
int ival;
#endif
+ static struct timeval lasttime;
+ static struct timeval interval = { .tv_sec = 1, .tv_usec = 0 };
KASSERT(p != NULL,
("%s() called without a process", __func__));
PROC_LOCK_ASSERT(p, MA_OWNED);
+
+ switch (cmd) {
+#if defined(COMPAT_FREEBSD5) || defined(COMPAT_FREEBSD4) || defined(COMPAT_43)
+ case _IOC(IOC_IN, 'p', 1, 0):
+#endif
+#ifdef COMPAT_FREEBSD6
+ case _IO('p', 1):
+#endif
+ case PIOCBIS:
+#if defined(COMPAT_FREEBSD5) || defined(COMPAT_FREEBSD4) || defined(COMPAT_43)
+ case _IOC(IOC_IN, 'p', 2, 0):
+#endif
+#ifdef COMPAT_FREEBSD6
+ case _IO('p', 2):
+#endif
+ case PIOCBIC:
+#if defined(COMPAT_FREEBSD5) || defined(COMPAT_FREEBSD4) || defined(COMPAT_43)
+ case _IOC(IOC_IN, 'p', 3, 0):
+#endif
+#ifdef COMPAT_FREEBSD6
+ case _IO('p', 3):
+#endif
+ case PIOCSFL:
+ case PIOCGFL:
+ case PIOCWAIT:
+ case PIOCSTATUS:
+#ifdef COMPAT_FREEBSD32
+ case PIOCWAIT32:
+ case PIOCSTATUS32:
+#endif
+#if defined(COMPAT_FREEBSD5) || defined(COMPAT_FREEBSD4) || defined(COMPAT_43)
+ case _IOC(IOC_IN, 'p', 5, 0):
+#endif
+#ifdef COMPAT_FREEBSD6
+ case _IO('p', 5):
+#endif
+ case PIOCCONT:
+ if (ratecheck(, ) != 0)
+ gone_in(13, "procfs-based process debugging");
+ break;
+ }
error = 0;
switch (cmd) {
Modified: head/sys/sys/pioctl.h
==
--- head/sys/sys/pioctl.h Tue Mar 17 18:17:32 2020(r359046)
+++ head/sys/sys/pioctl.h Tue Mar 17 18:44:03 2020(r359047)
@@ -41,6 +41,10 @@
#ifndef _SYS_PIOCTL_H
# define _SYS_PIOCTL_H
+#ifndef _KERNEL
+#warning " is deprecated, ptrace() should be used instead"
+#endif
+
# include
struct procfs_status {
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r358848 - in head/sys/amd64/vmm: intel io
On 3/10/20 9:53 AM, Michael Reifenberger wrote: > Author: mr > Date: Tue Mar 10 16:53:49 2020 > New Revision: 358848 > URL: https://svnweb.freebsd.org/changeset/base/358848 > > Log: > Untangle TPR shadowing and APIC virtualization. > This speeds up Windows guests tremendously. > > The patch does: > Add a new tuneable 'hw.vmm.vmx.use_tpr_shadowing' to disable TLP shadowing. > Also add 'hw.vmm.vmx.cap.tpr_shadowing' to be able to query if TPR > shadowing is used. > > Detach the initialization of TPR shadowing from the initialization of APIC > virtualization. > APIC virtualization still needs TPR shadowing, but not vice versa. > Any CPU that supports APIC virtualization should also support TPR shadowing. > > When TPR shadowing is used, the APIC page of each vCPU is written to the > VMCS_VIRTUAL_APIC field of the VMCS > so that the CPU can write directly to the page without intercept. > > On vm exit, vlapic_update_ppr() is called to update the PPR. > > Submitted by: Yamagi Burmeister > MFC after: 2 weeks > Differential Revision: https://reviews.freebsd.org/D22942 Reviewed by:grehan Also, it is still good form to wrap the commit logs to fit in 80 cols. -- John Baldwin ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r358789 - head/share/man/man7
On 3/8/20 2:30 PM, Ed Maste wrote: > Author: emaste > Date: Sun Mar 8 21:30:55 2020 > New Revision: 358789 > URL: https://svnweb.freebsd.org/changeset/base/358789 > > Log: > arch.7: remove note about GCC 4.2.1 removal date > > GCC 4.2.1 was removed in r358454. > > Sponsored by: The FreeBSD Foundation Can you update the platforms page on the website as well? -- John Baldwin ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r358687 - head/libexec/tftpd/tests
On 3/5/20 2:45 PM, Kyle Evans wrote:
> Author: kevans
> Date: Thu Mar 5 22:45:16 2020
> New Revision: 358687
> URL: https://svnweb.freebsd.org/changeset/base/358687
>
> Log:
> tftpd: tests: raise targeted cstd to c11
>
> r358556 added alignas() use to the functional tests, which isn't defined
> until C11. Raise the -std to C11 to fix the build under freebsd-gcc{6,9}.
>
> Reported by:mhorne, Jenkins/CI
Thanks for fixing!
--
John Baldwin
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r358561 - in head: . share/man/man5 share/man/man7 tools/build/options tools/tools/nanobsd/dhcpd tools/tools/nanobsd/embedded usr.bin usr.bin/calendar usr.bin/calendar/calendars usr.bi
On 3/2/20 3:37 PM, Conrad Meyer wrote: > Author: cem > Date: Mon Mar 2 23:37:47 2020 > New Revision: 358561 > URL: https://svnweb.freebsd.org/changeset/base/358561 > > Log: > Fix typo in r278616 > > FreeBSD isn't an encyclopedia. This isn't a helpful log message, and you didn't post this for review, etc. I actually use calendar in my .login, so it's not completely unused. It was a conscious decision, not a typo to keep calendar(1), and not one for you to unilaterally revisit without at least some discussion. I don't think we've had any calendar CVEs or that it's been a maintenance burden? -- John Baldwin ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r358556 - in head: libexec/tftpd libexec/tftpd/tests usr.bin/tftp
Author: jhb
Date: Mon Mar 2 22:19:30 2020
New Revision: 358556
URL: https://svnweb.freebsd.org/changeset/base/358556
Log:
Add support for the TFTP windowsize option described in RFC 7440.
The windowsize option permits multiple blocks to be transmitted
before the receiver sends an ACK improving throughput for larger
files.
Reviewed by: asomers
MFC after:2 weeks
Sponsored by: DARPA
Differential Revision:https://reviews.freebsd.org/D23836
Modified:
head/libexec/tftpd/tests/functional.c
head/libexec/tftpd/tftp-file.c
head/libexec/tftpd/tftp-file.h
head/libexec/tftpd/tftp-options.c
head/libexec/tftpd/tftp-options.h
head/libexec/tftpd/tftp-transfer.c
head/libexec/tftpd/tftp-utils.c
head/libexec/tftpd/tftp-utils.h
head/libexec/tftpd/tftpd.8
head/usr.bin/tftp/main.c
head/usr.bin/tftp/tftp.1
Modified: head/libexec/tftpd/tests/functional.c
==
--- head/libexec/tftpd/tests/functional.c Mon Mar 2 21:19:51 2020
(r358555)
+++ head/libexec/tftpd/tests/functional.c Mon Mar 2 22:19:30 2020
(r358556)
@@ -38,6 +38,7 @@ __FBSDID("$FreeBSD$");
#include
#include
#include
+#include
#include
#include
@@ -89,6 +90,13 @@ recv_ack(uint16_t blocknum)
RECV(hdr, NULL, 0);
}
+static void
+recv_oack(const char *options, size_t options_len)
+{
+ char hdr[] = {0, 6};
+ RECV(hdr, options, options_len);
+}
+
/*
* Receive a data packet from tftpd
* @param blocknumExpected block number to be received
@@ -159,6 +167,11 @@ send_ack(uint16_t blocknum)
}
+/*
+ * build an option string
+ */
+#define OPTION_STR(name, value)name "\000" value "\000"
+
/*
* send a read request to tftpd.
* @param filenamefilename as a string, absolute or relative
@@ -166,6 +179,11 @@ send_ack(uint16_t blocknum)
*/
#define SEND_RRQ(filename, mode) SEND_STR("\0\001" filename "\0" mode "\0")
+/*
+ * send a read request with options
+ */
+#define SEND_RRQ_OPT(filename, mode, options) SEND_STR("\0\001" filename "\0"
mode "\000" options)
+
/*
* send a write request to tftpd.
* @param filenamefilename as a string, absolute or relative
@@ -173,6 +191,11 @@ send_ack(uint16_t blocknum)
*/
#define SEND_WRQ(filename, mode) SEND_STR("\0\002" filename "\0" mode "\0")
+/*
+ * send a write request with options
+ */
+#define SEND_WRQ_OPT(filename, mode, options) SEND_STR("\0\002" filename "\0"
mode "\000" options)
+
/* Define a test case, for both IPv4 and IPv6 */
#define TFTPD_TC_DEFINE(name, head, ...) \
static void \
@@ -573,6 +596,32 @@ TFTPD_TC_DEFINE(rrq_medium,)
}
/*
+ * Read a medium file with a window size of 2.
+ */
+TFTPD_TC_DEFINE(rrq_medium_window,)
+{
+ int fd;
+ size_t i;
+ uint32_t contents[192];
+ char options[] = OPTION_STR("windowsize", "2");
+
+ for (i = 0; i < nitems(contents); i++)
+ contents[i] = i;
+
+ fd = open("medium.txt", O_RDWR | O_CREAT, 0644);
+ ATF_REQUIRE(fd >= 0);
+ write_all(fd, contents, sizeof(contents));
+ close(fd);
+
+ SEND_RRQ_OPT("medium.txt", "octet", OPTION_STR("windowsize", "2"));
+ recv_oack(options, sizeof(options) - 1);
+ send_ack(0);
+ recv_data(1, (const char*)[0], 512);
+ recv_data(2, (const char*)[128], 256);
+ send_ack(2);
+}
+
+/*
* Read a file in netascii format
*/
TFTPD_TC_DEFINE(rrq_netascii,)
@@ -652,6 +701,59 @@ TFTPD_TC_DEFINE(rrq_small,)
}
/*
+ * Read a file following the example in RFC 7440.
+ */
+TFTPD_TC_DEFINE(rrq_window_rfc7440,)
+{
+ int fd;
+ size_t i;
+ char options[] = OPTION_STR("windowsize", "4");
+ alignas(uint32_t) char contents[13 * 512 - 4];
+ uint32_t *u32p;
+
+ u32p = (uint32_t *)contents;
+ for (i = 0; i < sizeof(contents) / sizeof(uint32_t); i++)
+ u32p[i] = i;
+
+ fd = open("rfc7440.txt", O_RDWR | O_CREAT, 0644);
+ ATF_REQUIRE(fd >= 0);
+ write_all(fd, contents, sizeof(contents));
+ close(fd);
+
+ SEND_RRQ_OPT("rfc7440.txt", "octet", OPTION_STR("windowsize", "4"));
+ recv_oack(options, sizeof(options) - 1);
+ send_ack(0);
+ recv_data(1, [0 * 512], 512);
+ recv_data(2, [1 * 512], 512);
+ recv_data(3, [2 * 512], 512);
+ recv_data(4, [3 * 512], 512);
+ send_ack(4);
+ recv_data(5, [4 * 512], 512);
+ recv_data(6, [5 * 512], 512);
+ recv_data(7, [6 * 512], 512);
+ recv_data(8, [7 * 512], 512);
+
+ /* ACK 5 as if 6-8 were dropped. */
+ send_ack(5);
+ recv_data(6, [5 * 512], 512);
+ recv_data(7, [6 * 512], 512);
+ recv_data(8, [7 * 512], 512);
+ recv_data(9, [8 * 512], 512);
+ send_ack(9);
+ recv_data(10, [9 * 512], 512);
+ recv_data(11, [10 * 512], 512);
+ recv_data(12, [11 * 512], 512);
+ recv_data(13,
Re: svn commit: r358486 - head/sys/dev/bce
On 3/1/20 9:27 AM, Warner Losh wrote: > Author: imp > Date: Sun Mar 1 17:27:30 2020 > New Revision: 358486 > URL: https://svnweb.freebsd.org/changeset/base/358486 > > Log: > Remove all the compatibility hacks for systems that predate FreeBSD 8. Some > of > these look to be cut and pasted from other drivers since this driver was > committed to FreeBSD 7-current and MFC'd to FreeBSD 6. The ones for FreeBSD > 4 > and 5 likely never were working... You'd be surprised. I had to backport this to 4.x for Y!BSD, so I do suspect the 4.x bits worked fine. Good to clean out the cruft regardless. -- John Baldwin ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r358415 - in head/sys/dev/cxgbe: . tom
Author: jhb
Date: Fri Feb 28 00:42:27 2020
New Revision: 358415
URL: https://svnweb.freebsd.org/changeset/base/358415
Log:
Rename TOE TLS stats from [rt]x_tls_* to [rt]x_toe_tls_*.
This more clearly differentiates TLS records encrypted and decrypted
in TOE connections from those encrypted via NIC TLS.
MFC after:1 week
Sponsored by: Chelsio Communications
Modified:
head/sys/dev/cxgbe/adapter.h
head/sys/dev/cxgbe/t4_main.c
head/sys/dev/cxgbe/tom/t4_tls.c
Modified: head/sys/dev/cxgbe/adapter.h
==
--- head/sys/dev/cxgbe/adapter.hThu Feb 27 23:06:40 2020
(r358414)
+++ head/sys/dev/cxgbe/adapter.hFri Feb 28 00:42:27 2020
(r358415)
@@ -308,10 +308,10 @@ struct port_info {
struct port_stats stats;
u_int tnl_cong_drops;
u_int tx_parse_error;
- u_long tx_tls_records;
- u_long tx_tls_octets;
- u_long rx_tls_records;
- u_long rx_tls_octets;
+ u_long tx_toe_tls_records;
+ u_long tx_toe_tls_octets;
+ u_long rx_toe_tls_records;
+ u_long rx_toe_tls_octets;
struct callout tick;
};
Modified: head/sys/dev/cxgbe/t4_main.c
==
--- head/sys/dev/cxgbe/t4_main.cThu Feb 27 23:06:40 2020
(r358414)
+++ head/sys/dev/cxgbe/t4_main.cFri Feb 28 00:42:27 2020
(r358415)
@@ -7047,17 +7047,17 @@ cxgbe_sysctls(struct port_info *pi)
#undef SYSCTL_ADD_T4_PORTSTAT
- SYSCTL_ADD_ULONG(ctx, children, OID_AUTO, "tx_tls_records",
- CTLFLAG_RD, >tx_tls_records,
+ SYSCTL_ADD_ULONG(ctx, children, OID_AUTO, "tx_toe_tls_records",
+ CTLFLAG_RD, >tx_toe_tls_records,
"# of TOE TLS records transmitted");
- SYSCTL_ADD_ULONG(ctx, children, OID_AUTO, "tx_tls_octets",
- CTLFLAG_RD, >tx_tls_octets,
+ SYSCTL_ADD_ULONG(ctx, children, OID_AUTO, "tx_toe_tls_octets",
+ CTLFLAG_RD, >tx_toe_tls_octets,
"# of payload octets in transmitted TOE TLS records");
- SYSCTL_ADD_ULONG(ctx, children, OID_AUTO, "rx_tls_records",
- CTLFLAG_RD, >rx_tls_records,
+ SYSCTL_ADD_ULONG(ctx, children, OID_AUTO, "rx_toe_tls_records",
+ CTLFLAG_RD, >rx_toe_tls_records,
"# of TOE TLS records received");
- SYSCTL_ADD_ULONG(ctx, children, OID_AUTO, "rx_tls_octets",
- CTLFLAG_RD, >rx_tls_octets,
+ SYSCTL_ADD_ULONG(ctx, children, OID_AUTO, "rx_toe_tls_octets",
+ CTLFLAG_RD, >rx_toe_tls_octets,
"# of payload octets in received TOE TLS records");
}
Modified: head/sys/dev/cxgbe/tom/t4_tls.c
==
--- head/sys/dev/cxgbe/tom/t4_tls.c Thu Feb 27 23:06:40 2020
(r358414)
+++ head/sys/dev/cxgbe/tom/t4_tls.c Fri Feb 28 00:42:27 2020
(r358415)
@@ -1559,8 +1559,8 @@ t4_push_tls_records(struct adapter *sc, struct toepcb
}
toep->txsd_avail--;
- atomic_add_long(>vi->pi->tx_tls_records, 1);
- atomic_add_long(>vi->pi->tx_tls_octets, plen);
+ atomic_add_long(>vi->pi->tx_toe_tls_records, 1);
+ atomic_add_long(>vi->pi->tx_toe_tls_octets, plen);
t4_l2t_send(sc, wr, toep->l2te);
}
@@ -1862,8 +1862,8 @@ t4_push_ktls(struct adapter *sc, struct toepcb *toep,
}
toep->txsd_avail--;
- atomic_add_long(>vi->pi->tx_tls_records, 1);
- atomic_add_long(>vi->pi->tx_tls_octets, m->m_len);
+ atomic_add_long(>vi->pi->tx_toe_tls_records, 1);
+ atomic_add_long(>vi->pi->tx_toe_tls_octets, m->m_len);
t4_l2t_send(sc, wr, toep->l2te);
}
@@ -1899,7 +1899,7 @@ do_tls_data(struct sge_iq *iq, const struct rss_header
m_adj(m, sizeof(*cpl));
len = m->m_pkthdr.len;
- atomic_add_long(>vi->pi->rx_tls_octets, len);
+ atomic_add_long(>vi->pi->rx_toe_tls_octets, len);
KASSERT(len == G_CPL_TLS_DATA_LENGTH(be32toh(cpl->length_pkd)),
("%s: payload length mismatch", __func__));
@@ -1962,7 +1962,7 @@ do_rx_tls_cmp(struct sge_iq *iq, const struct rss_head
m_adj(m, sizeof(*cpl));
len = m->m_pkthdr.len;
- atomic_add_long(>vi->pi->rx_tls_records, 1);
+ atomic_add_long(>vi->pi->rx_toe_tls_records, 1);
KASSERT(len == G_CPL_RX_TLS_CMP_LENGTH(be32toh(cpl->pdulength_length)),
("%s: payload length mismatch", __func__));
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r358392 - head/share/man/man9
On 2/27/20 7:30 AM, Warner Losh wrote: > Author: imp > Date: Thu Feb 27 15:30:13 2020 > New Revision: 358392 > URL: https://svnweb.freebsd.org/changeset/base/358392 > > Log: > _Static_assert is to be preferred to CTASSERT. > > Document the existing prefernce that _Static_assert be used in preference > to the > old CTASSERT we used to use for compile time assertions. Actually, I think what we want to use is static_assert(). The intention in userland C is that _Static_assert() is an internal keyword and adds static_assert() as an alias, similar to defining alignas, etc. I think what we should do for the kernel is have define map static_assert to _Static_assert and replace existing _Static_assert usage with the proper spelling. -- John Baldwin ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r358394 - head/usr.sbin/bhyve
Author: jhb Date: Thu Feb 27 16:51:41 2020 New Revision: 358394 URL: https://svnweb.freebsd.org/changeset/base/358394 Log: Use stream_read() to read all 12 bytes of the RFB client version. read() can return a short read, whereas stream_read() waits until the full version string is read. Submitted by: Ka Ho Ng Reviewed by: grehan MFC after:1 week Differential Revision:https://reviews.freebsd.org/D23591 Modified: head/usr.sbin/bhyve/rfb.c Modified: head/usr.sbin/bhyve/rfb.c == --- head/usr.sbin/bhyve/rfb.c Thu Feb 27 15:34:30 2020(r358393) +++ head/usr.sbin/bhyve/rfb.c Thu Feb 27 16:51:41 2020(r358394) @@ -76,6 +76,7 @@ static int rfb_debug = 0; #defineDPRINTF(params) if (rfb_debug) PRINTLN params #defineWPRINTF(params) PRINTLN params +#define VERSION_LENGTH 12 #define AUTH_LENGTH16 #define PASSWD_LENGTH 8 @@ -769,7 +770,7 @@ rfb_handle(struct rfb_softc *rc, int cfd) stream_write(cfd, vbuf, strlen(vbuf)); /* 1b. Read client version */ - len = read(cfd, buf, sizeof(buf)); + len = stream_read(cfd, buf, VERSION_LENGTH); /* 2a. Send security type */ buf[0] = 1; ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r358346 - in head/libexec/rtld-elf: . rtld-libc
On 2/26/20 10:49 AM, Warner Losh wrote: > Author: imp > Date: Wed Feb 26 18:49:25 2020 > New Revision: 358346 > URL: https://svnweb.freebsd.org/changeset/base/358346 > > Log: > Remove sparc64 specific parts of rtld-elf. > > Modified: > head/libexec/rtld-elf/rtld-libc/Makefile.inc > head/libexec/rtld-elf/rtld.c > head/libexec/rtld-elf/rtld_tls.h libexec/rtld-elf/sparc64? -- John Baldwin ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r358247 - in head/sys: conf riscv/include
On 2/22/20 5:23 AM, Kristof Provost wrote: > Author: kp > Date: Sat Feb 22 13:23:27 2020 > New Revision: 358247 > URL: https://svnweb.freebsd.org/changeset/base/358247 > > Log: > riscv: Set MACHINE_ARCH correctly > > MACHINE_ARCH sets the hw.machine_arch sysctl in the kernel. In userspace > it sets MACHINE_ARCH in bmake, which bsd.cpu.mk uses to configure the > target ABI for ports. > > For riscv64sf builds (i.e. soft-float) that needs to be riscv64sf, but > the sysctl didn't reflect that. It is static. > > Set the define from the riscv makefile so that we correctly reflect our > actual build (i.e. riscv64 or riscv64sf), depending on what TARGET_ARCH > we were built with. > > That still doesn't satisfy userspace builds (e.g. bmake), so check if > we're building with a software-floating point toolchain there. That > check doesn't work in the kernel, because it never uses floating point. > > Reviewed by:philip (previous version), mhorne > Sponsored by: Axiado > Differential Revision: https://reviews.freebsd.org/D23741 I actually think this is kind of busted as the kernel is always the same, and we have the same problem on MIPS which also already does this wrong. I think instead we should treat soft-float as an alternate ABI (though very light weight as it doesn't need a separate syscall handler, etc.) and make the sysctl handler smart enough to return the right value based on the ELF header of the binary. I saw the later followup commit to add GENERICSF and I think that is possibly even farther down the wrong path. The kernel is _always_ soft-float in terms of the ABI it is built with. However, kernel with FPE support can run binaries with either ABI. It's true that on RISC-V there is not a nice way to detect if the FPU is present (I haven't looked at OpenSBI, but the SBI used with bbl doesn't provide misa to S-mode to know if it is present or not). You could just try to execute an FPU instruction during boot to see if you get an illegal instruction fault (subject to your SBI choosing to emulate that). -- John Baldwin ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r357891 - head/sys/dev/mlx5/mlx5_en
Author: jhb
Date: Thu Feb 13 23:04:11 2020
New Revision: 357891
URL: https://svnweb.freebsd.org/changeset/base/357891
Log:
Don't check the auth algorithm for GCM.
The upstream OpenSSL changes only set the cipher for GCM since the
authentication is redundant, and changes to OCF will soon remove the
GCM authentication algorithm constants entirely for the same reason.
In addition, ktls_create_session() already validates these fields and
wouldn't pass down an invalid auth_algorithm value to any drivers or
ktls backends.
Reviewed by: hselasky
Sponsored by: Chelsio Communications
Differential Revision:https://reviews.freebsd.org/D23671
Modified:
head/sys/dev/mlx5/mlx5_en/mlx5_en_hw_tls.c
Modified: head/sys/dev/mlx5/mlx5_en/mlx5_en_hw_tls.c
==
--- head/sys/dev/mlx5/mlx5_en/mlx5_en_hw_tls.c Thu Feb 13 22:55:45 2020
(r357890)
+++ head/sys/dev/mlx5/mlx5_en/mlx5_en_hw_tls.c Thu Feb 13 23:04:11 2020
(r357891)
@@ -339,10 +339,6 @@ mlx5e_tls_snd_tag_alloc(struct ifnet *ifp,
case CRYPTO_AES_NIST_GCM_16:
switch (en->cipher_key_len) {
case 128 / 8:
- if (en->auth_algorithm != CRYPTO_AES_128_NIST_GMAC) {
- error = EINVAL;
- goto failure;
- }
if (en->tls_vminor == TLS_MINOR_VER_TWO) {
if (MLX5_CAP_TLS(priv->mdev,
tls_1_2_aes_gcm_128) == 0) {
error = EPROTONOSUPPORT;
@@ -360,10 +356,6 @@ mlx5e_tls_snd_tag_alloc(struct ifnet *ifp,
break;
case 256 / 8:
- if (en->auth_algorithm != CRYPTO_AES_256_NIST_GMAC) {
- error = EINVAL;
- goto failure;
- }
if (en->tls_vminor == TLS_MINOR_VER_TWO) {
if (MLX5_CAP_TLS(priv->mdev,
tls_1_2_aes_gcm_256) == 0) {
error = EPROTONOSUPPORT;
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r357890 - in head/sys/dev/cxgbe: . crypto
Author: jhb
Date: Thu Feb 13 22:55:45 2020
New Revision: 357890
URL: https://svnweb.freebsd.org/changeset/base/357890
Log:
Remove the per-TXQ tls_wrs stat.
It duplicated the kern_tls_records stat and was not conditional on NIC
TLS being enabled.
Reviewed by: np
Sponsored by: Chelsio Communications
Differential Revision:https://reviews.freebsd.org/D23670
Modified:
head/sys/dev/cxgbe/adapter.h
head/sys/dev/cxgbe/crypto/t4_kern_tls.c
head/sys/dev/cxgbe/t4_main.c
head/sys/dev/cxgbe/t4_sge.c
Modified: head/sys/dev/cxgbe/adapter.h
==
--- head/sys/dev/cxgbe/adapter.hThu Feb 13 22:22:55 2020
(r357889)
+++ head/sys/dev/cxgbe/adapter.hThu Feb 13 22:55:45 2020
(r357890)
@@ -574,7 +574,6 @@ struct sge_txq {
uint64_t txpkts0_pkts; /* # of frames in type0 coalesced tx WRs */
uint64_t txpkts1_pkts; /* # of frames in type1 coalesced tx WRs */
uint64_t raw_wrs; /* # of raw work requests (alloc_wr_mbuf) */
- uint64_t tls_wrs; /* # of TLS work requests */
uint64_t kern_tls_records;
uint64_t kern_tls_short;
Modified: head/sys/dev/cxgbe/crypto/t4_kern_tls.c
==
--- head/sys/dev/cxgbe/crypto/t4_kern_tls.c Thu Feb 13 22:22:55 2020
(r357889)
+++ head/sys/dev/cxgbe/crypto/t4_kern_tls.c Thu Feb 13 22:55:45 2020
(r357890)
@@ -2082,7 +2082,6 @@ ktls_write_tls_wr(struct tlspcb *tlsp, struct sge_txq
ndesc += howmany(wr_len, EQ_ESIZE);
MPASS(ndesc <= available);
- txq->tls_wrs++;
txq->kern_tls_records++;
txq->kern_tls_octets += tlen - mtod(m_tls, vm_offset_t);
Modified: head/sys/dev/cxgbe/t4_main.c
==
--- head/sys/dev/cxgbe/t4_main.cThu Feb 13 22:22:55 2020
(r357889)
+++ head/sys/dev/cxgbe/t4_main.cThu Feb 13 22:55:45 2020
(r357890)
@@ -10359,7 +10359,6 @@ clear_stats(struct adapter *sc, u_int port_id)
txq->txpkts0_pkts = 0;
txq->txpkts1_pkts = 0;
txq->raw_wrs = 0;
- txq->tls_wrs = 0;
txq->kern_tls_records = 0;
txq->kern_tls_short = 0;
txq->kern_tls_partial = 0;
Modified: head/sys/dev/cxgbe/t4_sge.c
==
--- head/sys/dev/cxgbe/t4_sge.c Thu Feb 13 22:22:55 2020(r357889)
+++ head/sys/dev/cxgbe/t4_sge.c Thu Feb 13 22:55:45 2020(r357890)
@@ -4204,8 +4204,6 @@ alloc_txq(struct vi_info *vi, struct sge_txq *txq, int
"# of frames tx'd using type1 txpkts work requests");
SYSCTL_ADD_UQUAD(>ctx, children, OID_AUTO, "raw_wrs", CTLFLAG_RD,
>raw_wrs, "# of raw work requests (non-packets)");
- SYSCTL_ADD_UQUAD(>ctx, children, OID_AUTO, "tls_wrs", CTLFLAG_RD,
- >tls_wrs, "# of TLS work requests (TLS records)");
#ifdef KERN_TLS
if (sc->flags & KERN_TLS_OK) {
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r357361 - in head/sys: kern sys ufs/ufs vm
On 2/10/20 11:54 AM, Mateusz Guzik wrote: > On 2/3/20, John Baldwin wrote: >> On 1/31/20 10:46 PM, Mateusz Guzik wrote: >>> Author: mjg >>> Date: Sat Feb 1 06:46:55 2020 >>> New Revision: 357361 >>> URL: https://svnweb.freebsd.org/changeset/base/357361 >>> >>> Log: >>> vfs: replace VOP_MARKATIME with VOP_MMAPPED >>> >>> The routine is only provided by ufs and is only used on mmap and exec. >>> >>> Reviewed by: kib >>> Differential Revision:https://reviews.freebsd.org/D23422 >>> >>> Modified: >>> head/sys/kern/kern_exec.c >>> head/sys/kern/vfs_subr.c >>> head/sys/kern/vnode_if.src >>> head/sys/sys/vnode.h >>> head/sys/ufs/ufs/ufs_vnops.c >>> head/sys/vm/vm_mmap.c >>> >>> Modified: head/sys/ufs/ufs/ufs_vnops.c >>> == >>> --- head/sys/ufs/ufs/ufs_vnops.cSat Feb 1 06:41:44 2020 >>> (r357360) >>> +++ head/sys/ufs/ufs/ufs_vnops.cSat Feb 1 06:46:55 2020 >>> (r357361) >>> @@ -108,7 +108,7 @@ static vop_getattr_tufs_getattr; >>> static vop_ioctl_t ufs_ioctl; >>> static vop_link_t ufs_link; >>> static int ufs_makeinode(int mode, struct vnode *, struct vnode **, >>> struct componentname *, const char *); >>> -static vop_markatime_t ufs_markatime; >>> +static vop_mmapped_t ufs_mmapped; >>> static vop_mkdir_t ufs_mkdir; >>> static vop_mknod_t ufs_mknod; >>> static vop_open_t ufs_open; >>> @@ -676,19 +676,22 @@ out: >>> } >>> #endif /* UFS_ACL */ >>> >>> -/* >>> - * Mark this file's access time for update for vfs_mark_atime(). This >>> - * is called from execve() and mmap(). >>> - */ >> >> Why remove this comment rather than update it? It is largely still >> true and explains the purpose of the VOP (update the atime) which is >> now no longer obvious from the name. >> > > I don't think a fs-specific implementation of a VOP is the right place to > state where it is called from. I would argue the name could be better as > the execve bit is definitely not obvious, but interested parties can > always grep. This (always grep) is why comments matter. If we wanted people to just use grep and always UTSL, we wouldn't bother having any comments at all. One of the purposes of comments is to allow a human reader to understand the code in context without needing several different windows open to piece together what is happening. In particular, the comment gives a better hint as to _why_ we are updating the atime. The old name did a better job of this than the new one which is more bland, but presumably you have future changes to add that are beyond just changing the atime and that is why you renamed it? -- John Baldwin ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r357695 - in head: sys/kern sys/sys usr.bin/procstat
On 2/10/20 11:15 AM, Konstantin Belousov wrote:
> On Mon, Feb 10, 2020 at 10:11:43AM -0800, John Baldwin wrote:
>> On 2/9/20 4:10 AM, Konstantin Belousov wrote:
>>> Author: kib
>>> Date: Sun Feb 9 12:10:37 2020
>>> New Revision: 357695
>>> URL: https://svnweb.freebsd.org/changeset/base/357695
>>>
>>> Log:
>>> Add AT_BSDFLAGS auxv entry.
>>>
>>> The intent is to provide bsd-specific flags relevant to interpreter
>>> and C runtime. I did not want to reuse AT_FLAGS which is common ELF
>>> auxv entry.
>>>
>>> Use bsdflags to report kernel support for sigfastblock(2). This
>>> allows rtld and libthr to safely infer the syscall presence without
>>> SIGSYS. The tunable kern.elf{32,64}.sigfastblock blocks reporting.
>>>
>>> Tested by:pho
>>> Disscussed with: cem, emaste, jilles
>>> Sponsored by: The FreeBSD Foundation
>>> Differential revision:https://reviews.freebsd.org/D12773
>>
>> I find adding a new auxv type curious. The MIPS ABI doc says that
>> "bits under the 0xff00 mask are reserved for system semantics".
>> The powerpc and x86-64 docs don't define any bits at all. In
>> practice I think we are free to use AT_FLAGS however we wish as no
>> use cases of "standard" bits have arisen since AT_FLAGS was first
>> defined.
>
> So you would prefer to have me used AT_FLAGS for sigfastblock indicator ?
> I am feeling uncomfortable doing that.
>
> My reasoning, to reformulate it from what I wrote in the commit message,
> is to not pollute neither compilation nor ABI namespace for bsd-specific
> flags. AT_FLAGS was not touched by anybody and I do not want to open
> it for use, since ABI group my finally find some use for it.
I would bet money the ABI group will never use AT_FLAGS since they haven't
found a use yet. That said, adding a new auxv vector isn't the end of the
world. I'll work on a patch to GDB when I get some spare time.
--
John Baldwin
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r357740 - head/usr.sbin/ntp
On 2/10/20 9:17 AM, Warner Losh wrote: > Author: imp > Date: Mon Feb 10 17:17:03 2020 > New Revision: 357740 > URL: https://svnweb.freebsd.org/changeset/base/357740 > > Log: > Refresh architecture list with latest: > > Remove: sparc, sparc64 (twice), ia64, alpha > Add: riscv64 > > Remove now-useless sparc64 ifdef > > Reivewed by: cy@, bcr@ > Differential Revision: https://reviews.freebsd.org/D23593 > > Modified: > head/usr.sbin/ntp/config.h > > Modified: head/usr.sbin/ntp/config.h > == > --- head/usr.sbin/ntp/config.hMon Feb 10 17:16:59 2020 > (r357739) > +++ head/usr.sbin/ntp/config.hMon Feb 10 17:17:03 2020 > (r357740) > @@ -1578,11 +1578,7 @@ > /* #undef STRERROR_R_CHAR_P */ > > /* canonical system (cpu-vendor-os) of where we should run */ > -#if defined(__alpha__) > -#define STR_SYSTEM "alpha-undermydesk-freebsd" > -#elif defined(__sparc64__) > -#define STR_SYSTEM "sparc64-undermydesk-freebsd" > -#elif defined(__amd64__) > +#if defined(__amd64__) > #define STR_SYSTEM "amd64-undermydesk-freebsd" > #elif defined(__powerpc64__) > #define STR_SYSTEM "powerpc64-undermydesk-freebsd" > @@ -1596,12 +1592,8 @@ > #define STR_SYSTEM "arm64-undermydesk-freebsd" > #elif defined(__arm__) > #define STR_SYSTEM "arm-undermydesk-freebsd" > -#elif defined(__sparc64__) > -#define STR_SYSTEM "sparc64-undermydesk-freebsd" > -#elif defined(__sparc__) > -#define STR_SYSTEM "sparc-undermydesk-freebsd" > -#elif defined(__ia64__) > -#define STR_SYSTEM "ia64-undermydesk-freebsd" > +#elif defined(__riscv__) > +#define STR_SYSTEM "riscv64-undermydesk-freebsd" This should be __riscv (see arch(7)) -- John Baldwin ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r357703 - head/share/mk
On 2/9/20 10:53 AM, Kyle Evans wrote: > Author: kevans > Date: Sun Feb 9 18:53:53 2020 > New Revision: 357703 > URL: https://svnweb.freebsd.org/changeset/base/357703 > > Log: > mips: mark GOOGLETEST broken, due to no fault of its own > > As explained in the comment; GOOGLETEST cannot currently be compiled on any > mips variant at the moment due to the cross toolchain seemingly using the > wrong spec and not pulling in libgcc. We'll be fine when llvm 10 lands, at > which point this should be reverted most expeditiously. Hmm, I thought you were going to do this for any external GCC? It's also broken on amd64, etc. as well when using external GCC. -- John Baldwin ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r357695 - in head: sys/kern sys/sys usr.bin/procstat
On 2/9/20 4:10 AM, Konstantin Belousov wrote:
> Author: kib
> Date: Sun Feb 9 12:10:37 2020
> New Revision: 357695
> URL: https://svnweb.freebsd.org/changeset/base/357695
>
> Log:
> Add AT_BSDFLAGS auxv entry.
>
> The intent is to provide bsd-specific flags relevant to interpreter
> and C runtime. I did not want to reuse AT_FLAGS which is common ELF
> auxv entry.
>
> Use bsdflags to report kernel support for sigfastblock(2). This
> allows rtld and libthr to safely infer the syscall presence without
> SIGSYS. The tunable kern.elf{32,64}.sigfastblock blocks reporting.
>
> Tested by: pho
> Disscussed with:cem, emaste, jilles
> Sponsored by: The FreeBSD Foundation
> Differential revision: https://reviews.freebsd.org/D12773
I find adding a new auxv type curious. The MIPS ABI doc says that
"bits under the 0xff00 mask are reserved for system semantics".
The powerpc and x86-64 docs don't define any bits at all. In
practice I think we are free to use AT_FLAGS however we wish as no
use cases of "standard" bits have arisen since AT_FLAGS was first
defined.
--
John Baldwin
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r357643 - head/lib/libc/riscv/gen
Author: jhb
Date: Thu Feb 6 21:46:15 2020
New Revision: 357643
URL: https://svnweb.freebsd.org/changeset/base/357643
Log:
Tidy the _set_tp function for RISC-V.
- Use a constant for the offset instead of a magic number.
- Use an addi instruction that writes to tp directly instead of a mv
that writes the result of a compiler-generated addi.
Reviewed by: mhorne
MFC after:1 week
Sponsored by: DARPA
Differential Revision:https://reviews.freebsd.org/D23521
Modified:
head/lib/libc/riscv/gen/_set_tp.c
Modified: head/lib/libc/riscv/gen/_set_tp.c
==
--- head/lib/libc/riscv/gen/_set_tp.c Thu Feb 6 21:01:19 2020
(r357642)
+++ head/lib/libc/riscv/gen/_set_tp.c Thu Feb 6 21:46:15 2020
(r357643)
@@ -38,13 +38,14 @@ __FBSDID("$FreeBSD$");
#include
#include
-#include
-
#include
+/* NB: size of 'struct tcb'. */
+#defineTP_OFFSET (sizeof(void *) * 2)
+
void
_set_tp(void *tp)
{
- __asm __volatile("mv tp, %0" :: "r"((char*)tp + 0x10));
+ __asm __volatile("addi tp, %0, %1" :: "r" (tp), "I" (TP_OFFSET));
}
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r357632 - head/sys/riscv/riscv
Author: jhb
Date: Thu Feb 6 18:04:45 2020
New Revision: 357632
URL: https://svnweb.freebsd.org/changeset/base/357632
Log:
Use the context created in makectx() for stack traces.
Always use the kdb_thr_ctx() for db_trace_thread() as on other
architectures. Initialize pcb_ra to be the sepc from the saved
trapframe rather than the saved ra to avoid skipping a frame.
Reviewed by: mhorne, br
MFC after:1 week
Sponsored by: DARPA
Differential Revision:https://reviews.freebsd.org/D23513
Modified:
head/sys/riscv/riscv/db_trace.c
head/sys/riscv/riscv/machdep.c
Modified: head/sys/riscv/riscv/db_trace.c
==
--- head/sys/riscv/riscv/db_trace.c Thu Feb 6 18:04:15 2020
(r357631)
+++ head/sys/riscv/riscv/db_trace.c Thu Feb 6 18:04:45 2020
(r357632)
@@ -130,15 +130,12 @@ db_trace_thread(struct thread *thr, int count)
struct unwind_state frame;
struct pcb *ctx;
- if (thr != curthread) {
- ctx = kdb_thr_ctx(thr);
+ ctx = kdb_thr_ctx(thr);
- frame.sp = (uint64_t)ctx->pcb_sp;
- frame.fp = (uint64_t)ctx->pcb_s[0];
- frame.pc = (uint64_t)ctx->pcb_ra;
- db_stack_trace_cmd();
- } else
- db_trace_self();
+ frame.sp = (uint64_t)ctx->pcb_sp;
+ frame.fp = (uint64_t)ctx->pcb_s[0];
+ frame.pc = (uint64_t)ctx->pcb_ra;
+ db_stack_trace_cmd();
return (0);
}
Modified: head/sys/riscv/riscv/machdep.c
==
--- head/sys/riscv/riscv/machdep.c Thu Feb 6 18:04:15 2020
(r357631)
+++ head/sys/riscv/riscv/machdep.c Thu Feb 6 18:04:45 2020
(r357632)
@@ -571,7 +571,7 @@ makectx(struct trapframe *tf, struct pcb *pcb)
memcpy(pcb->pcb_s, tf->tf_s, sizeof(tf->tf_s));
- pcb->pcb_ra = tf->tf_ra;
+ pcb->pcb_ra = tf->tf_sepc;
pcb->pcb_sp = tf->tf_sp;
pcb->pcb_gp = tf->tf_gp;
pcb->pcb_tp = tf->tf_tp;
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r357630 - head/sys/riscv/riscv
Author: jhb
Date: Thu Feb 6 18:02:38 2020
New Revision: 357630
URL: https://svnweb.freebsd.org/changeset/base/357630
Log:
Fix DDB to unwind across exception frames.
While here, add an extra line of information for exceptions and
interrupts and compress the per-frame line down to one line to match
other architectures.
Reviewed by: mhorne, br
MFC after:1 week
Sponsored by: DARPA
Differential Revision:https://reviews.freebsd.org/D23508
Modified:
head/sys/riscv/riscv/db_trace.c
Modified: head/sys/riscv/riscv/db_trace.c
==
--- head/sys/riscv/riscv/db_trace.c Thu Feb 6 18:00:50 2020
(r357629)
+++ head/sys/riscv/riscv/db_trace.c Thu Feb 6 18:02:38 2020
(r357630)
@@ -2,6 +2,7 @@
* Copyright (c) 2015 The FreeBSD Foundation
* Copyright (c) 2016 Ruslan Bukin
* All rights reserved.
+ * Copyright (c) 2020 John Baldwin
*
* Portions of this software were developed by Semihalf under
* the sponsorship of the FreeBSD Foundation.
@@ -38,15 +39,18 @@
#include
__FBSDID("$FreeBSD$");
+
#include
-#include
#include
-#include
+#include
+
#include
#include
+#include
#include
#include
+#include
void
db_md_list_watchpoints()
@@ -80,9 +84,6 @@ db_stack_trace_cmd(struct unwind_state *frame)
while (1) {
pc = frame->pc;
- if (unwind_frame(frame) < 0)
- break;
-
sym = db_search_symbol(pc, DB_STGY_ANY, );
if (sym == C_DB_SYM_NULL) {
value = 0;
@@ -94,11 +95,32 @@ db_stack_trace_cmd(struct unwind_state *frame)
db_printsym(frame->pc, DB_STGY_PROC);
db_printf("\n");
- db_printf("\t pc = 0x%016lx ra = 0x%016lx\n",
- pc, frame->pc);
- db_printf("\t sp = 0x%016lx fp = 0x%016lx\n",
- frame->sp, frame->fp);
- db_printf("\n");
+ if (strcmp(name, "cpu_exception_handler_supervisor") == 0 ||
+ strcmp(name, "cpu_exception_handler_user") == 0) {
+ struct trapframe *tf;
+
+ tf = (struct trapframe *)(uintptr_t)frame->sp;
+
+ if (tf->tf_scause & EXCP_INTR)
+ db_printf("--- interrupt %ld\n",
+ tf->tf_scause & EXCP_MASK);
+ else
+ db_printf("--- exception %ld, tval = %#lx\n",
+ tf->tf_scause & EXCP_MASK,
+ tf->tf_stval);
+ frame->sp = (uint64_t)tf->tf_sp;
+ frame->fp = (uint64_t)tf->tf_s[0];
+ frame->pc = (uint64_t)tf->tf_sepc;
+ if (!INKERNEL(frame->fp))
+ break;
+ continue;
+ }
+
+ if (strcmp(name, "fork_trampoline") == 0)
+ break;
+
+ if (unwind_frame(frame) < 0)
+ break;
}
}
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r357595 - head/sys/riscv/include
Author: jhb Date: Wed Feb 5 20:34:22 2020 New Revision: 357595 URL: https://svnweb.freebsd.org/changeset/base/357595 Log: Fix EXCP_MASK to include all relevant bits from scause. While cause codes higher than 16 are reserved, the exception code field of the register is defined to be all bits but the upper-most bit. Reviewed by: mhorne MFC after:1 week Sponsored by: DARPA Differential Revision:https://reviews.freebsd.org/D23510 Modified: head/sys/riscv/include/riscvreg.h Modified: head/sys/riscv/include/riscvreg.h == --- head/sys/riscv/include/riscvreg.h Wed Feb 5 20:32:37 2020 (r357594) +++ head/sys/riscv/include/riscvreg.h Wed Feb 5 20:34:22 2020 (r357595) @@ -37,8 +37,7 @@ #ifndef _MACHINE_RISCVREG_H_ #define_MACHINE_RISCVREG_H_ -#defineEXCP_SHIFT 0 -#defineEXCP_MASK (0xf << EXCP_SHIFT) +#defineEXCP_MASK (~EXCP_INTR) #defineEXCP_MISALIGNED_FETCH 0 #defineEXCP_FAULT_FETCH1 #defineEXCP_ILLEGAL_INSTRUCTION2 ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r357594 - head/sys/riscv/riscv
Author: jhb
Date: Wed Feb 5 20:32:37 2020
New Revision: 357594
URL: https://svnweb.freebsd.org/changeset/base/357594
Log:
Use csr_read() to read sstatus instead of inline assembly.
While here, remove a local variable to avoid the CSR read in non-debug
kernels.
Reviewed by: mhorne
MFC after:1 week
Sponsored by: DARPA
Differential Revision:https://reviews.freebsd.org/D23511
Modified:
head/sys/riscv/riscv/trap.c
Modified: head/sys/riscv/riscv/trap.c
==
--- head/sys/riscv/riscv/trap.c Wed Feb 5 20:11:08 2020(r357593)
+++ head/sys/riscv/riscv/trap.c Wed Feb 5 20:32:37 2020(r357594)
@@ -245,12 +245,10 @@ void
do_trap_supervisor(struct trapframe *frame)
{
uint64_t exception;
- uint64_t sstatus;
/* Ensure we came from supervisor mode, interrupts disabled */
- __asm __volatile("csrr %0, sstatus" : "=" (sstatus));
- KASSERT((sstatus & (SSTATUS_SPP | SSTATUS_SIE)) == SSTATUS_SPP,
- ("We must came from S mode with interrupts disabled"));
+ KASSERT((csr_read(sstatus) & (SSTATUS_SPP | SSTATUS_SIE)) ==
+ SSTATUS_SPP, ("Came from S mode with interrupts enabled"));
exception = (frame->tf_scause & EXCP_MASK);
if (frame->tf_scause & EXCP_INTR) {
@@ -305,7 +303,6 @@ do_trap_user(struct trapframe *frame)
{
uint64_t exception;
struct thread *td;
- uint64_t sstatus;
struct pcb *pcb;
td = curthread;
@@ -313,9 +310,8 @@ do_trap_user(struct trapframe *frame)
pcb = td->td_pcb;
/* Ensure we came from usermode, interrupts disabled */
- __asm __volatile("csrr %0, sstatus" : "=" (sstatus));
- KASSERT((sstatus & (SSTATUS_SPP | SSTATUS_SIE)) == 0,
- ("We must came from U mode with interrupts disabled"));
+ KASSERT((csr_read(sstatus) & (SSTATUS_SPP | SSTATUS_SIE)) == 0,
+ ("Came from U mode with interrupts enabled"));
exception = (frame->tf_scause & EXCP_MASK);
if (frame->tf_scause & EXCP_INTR) {
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r357593 - head/sys/riscv/riscv
Author: jhb
Date: Wed Feb 5 20:11:08 2020
New Revision: 357593
URL: https://svnweb.freebsd.org/changeset/base/357593
Log:
Remove stale workaround for the htif console.
In practice this discarded all characters entered at the DDB prompt.
Reviewed by: br
MFC after:1 week
Sponsored by: DARPA
Differential Revision:https://reviews.freebsd.org/D23509
Modified:
head/sys/riscv/riscv/riscv_console.c
Modified: head/sys/riscv/riscv/riscv_console.c
==
--- head/sys/riscv/riscv/riscv_console.cWed Feb 5 20:08:01 2020
(r357592)
+++ head/sys/riscv/riscv/riscv_console.cWed Feb 5 20:11:08 2020
(r357593)
@@ -206,20 +206,6 @@ riscv_cngetc(struct consdev *cp)
{
int ch;
-#if defined(KDB)
- /*
-* RISCVTODO: BBL polls for console data on timer interrupt,
-* but interrupts are turned off in KDB.
-* So we currently do not have console in KDB.
-*/
- if (kdb_active) {
- ch = sbi_console_getchar();
- while (ch) {
- ch = sbi_console_getchar();
- }
- }
-#endif
-
ch = sbi_console_getchar();
if (ch > 0 && ch < 0xff) {
#if defined(KDB)
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r357591 - head/sys/riscv/include
Author: jhb
Date: Wed Feb 5 20:06:35 2020
New Revision: 357591
URL: https://svnweb.freebsd.org/changeset/base/357591
Log:
Read the breakpoint instruction to determine its length in BKPT_SKIP.
This fixes continuing from debug.kdb.enter=1 after enabling the use of
compressed instructions since the compiler can emit the two byte
c.ebreak instead of the 4 byte ebreak.
Reviewed by: br
MFC after:1 week
Sponsored by: DARPA
Differential Revision:https://reviews.freebsd.org/D23507
Modified:
head/sys/riscv/include/db_machdep.h
Modified: head/sys/riscv/include/db_machdep.h
==
--- head/sys/riscv/include/db_machdep.h Wed Feb 5 19:39:48 2020
(r357590)
+++ head/sys/riscv/include/db_machdep.h Wed Feb 5 20:06:35 2020
(r357591)
@@ -53,8 +53,14 @@ typedef long db_expr_t;
#defineBKPT_SIZE (INSN_SIZE)
#defineBKPT_SET(inst) (BKPT_INST)
-#defineBKPT_SKIP do { \
- kdb_frame->tf_sepc += BKPT_SIZE;\
+#defineBKPT_SKIP do {
\
+ uint32_t _instr;\
+ \
+ _instr = db_get_value(PC_REGS(), sizeof(uint32_t), FALSE); \
+ if ((_instr & 0x3) == 0x3) \
+ kdb_frame->tf_sepc += 4;/* ebreak */\
+ else\
+ kdb_frame->tf_sepc += 2;/* c.ebreak */ \
} while (0)
#definedb_clear_single_stepkdb_cpu_clear_singlestep
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r357361 - in head/sys: kern sys ufs/ufs vm
On 1/31/20 10:46 PM, Mateusz Guzik wrote: > Author: mjg > Date: Sat Feb 1 06:46:55 2020 > New Revision: 357361 > URL: https://svnweb.freebsd.org/changeset/base/357361 > > Log: > vfs: replace VOP_MARKATIME with VOP_MMAPPED > > The routine is only provided by ufs and is only used on mmap and exec. > > Reviewed by:kib > Differential Revision: https://reviews.freebsd.org/D23422 > > Modified: > head/sys/kern/kern_exec.c > head/sys/kern/vfs_subr.c > head/sys/kern/vnode_if.src > head/sys/sys/vnode.h > head/sys/ufs/ufs/ufs_vnops.c > head/sys/vm/vm_mmap.c > > Modified: head/sys/ufs/ufs/ufs_vnops.c > == > --- head/sys/ufs/ufs/ufs_vnops.c Sat Feb 1 06:41:44 2020 > (r357360) > +++ head/sys/ufs/ufs/ufs_vnops.c Sat Feb 1 06:46:55 2020 > (r357361) > @@ -108,7 +108,7 @@ static vop_getattr_t ufs_getattr; > static vop_ioctl_t ufs_ioctl; > static vop_link_tufs_link; > static int ufs_makeinode(int mode, struct vnode *, struct vnode **, struct > componentname *, const char *); > -static vop_markatime_t ufs_markatime; > +static vop_mmapped_t ufs_mmapped; > static vop_mkdir_t ufs_mkdir; > static vop_mknod_t ufs_mknod; > static vop_open_tufs_open; > @@ -676,19 +676,22 @@ out: > } > #endif /* UFS_ACL */ > > -/* > - * Mark this file's access time for update for vfs_mark_atime(). This > - * is called from execve() and mmap(). > - */ Why remove this comment rather than update it? It is largely still true and explains the purpose of the VOP (update the atime) which is now no longer obvious from the name. -- John Baldwin ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r357407 - head
On 2/2/20 3:37 AM, Warner Losh wrote:
> Author: imp
> Date: Sun Feb 2 11:37:27 2020
> New Revision: 357407
> URL: https://svnweb.freebsd.org/changeset/base/357407
>
> Log:
> Per the firm plan, start to remove sparc64
>
> The sparc64 architecture is being removed from FreeBSD 13, starting
> now. This removes it from the top level only. It is the only
> architecture that didn't see substantial work after the call to get
> things working with the external toolchain.
>
> Modified:
> head/Makefile
> head/Makefile.inc1
>
> Modified: head/Makefile.inc1
> ==
> --- head/Makefile.inc1Sun Feb 2 09:38:40 2020(r357406)
> +++ head/Makefile.inc1Sun Feb 2 11:37:27 2020(r357407)
> @@ -2915,14 +2912,10 @@ _cddl_lib_libctf= cddl/lib/libctf
> _cddl_lib= cddl/lib
> cddl/lib/libctf__L: lib/libz__L
> .endif
> -# cddl/lib/libdtrace requires lib/libproc and lib/librtld_db; it's only built
> -# on select architectures though (see cddl/lib/Makefile)
The comment should stay, probably as:
# cddl/lib/libdtrace requires lib/libproc and lib/librtld_db
> -.if ${MACHINE_CPUARCH} != "sparc64"
> _prebuild_libs+= lib/libprocstat lib/libproc lib/librtld_db
> lib/libprocstat__L: lib/libelf__L lib/libkvm__L lib/libutil__L
> lib/libproc__L: lib/libprocstat__L
> lib/librtld_db__L: lib/libprocstat__L
> -.endif
--
John Baldwin
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r357337 - head/sys/riscv/include
On 1/31/20 1:17 PM, Conrad Meyer wrote: > Hi John, > > Isn't the 32-bit MSTATUS_SD and SSTATUS_SD also UB without explicit > unsigned suffix? Possibly, but I doubt we will have a 32-bit RISC-V port. I'd probably prefer to just use explicit hex values (0x8000, etc.) since the compiler will always get those correct. -- John Baldwin ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r357344 - head/sys/riscv/riscv
Author: jhb
Date: Fri Jan 31 19:00:48 2020
New Revision: 357344
URL: https://svnweb.freebsd.org/changeset/base/357344
Log:
Add stricter checks on user changes to SSTATUS.
Rather than trying to blacklist which bits userland can't write to via
sigreturn() or setcontext(), only permit changes to whitelisted bits.
- Permit arbitrary writes to bits in the user-writable USTATUS
register that shadows SSTATUS.
- Ignore changes in write-only bits maintained by the CPU.
- Ignore the user-supplied value of the FS field used to track
floating point state and instead set it to a value matching the
actions taken by set_fpcontext().
Discussed with: mhorne
MFC after:2 weeks
Sponsored by: DARPA
Differential Revision:https://reviews.freebsd.org/D23338
Modified:
head/sys/riscv/riscv/machdep.c
Modified: head/sys/riscv/riscv/machdep.c
==
--- head/sys/riscv/riscv/machdep.c Fri Jan 31 18:55:21 2020
(r357343)
+++ head/sys/riscv/riscv/machdep.c Fri Jan 31 19:00:48 2020
(r357344)
@@ -368,11 +368,16 @@ set_mcontext(struct thread *td, mcontext_t *mcp)
tf = td->td_frame;
/*
-* Make sure the processor mode has not been tampered with and
-* interrupts have not been disabled.
-* Supervisor interrupts in user mode are always enabled.
+* Permit changes to the USTATUS bits of SSTATUS.
+*
+* Ignore writes to read-only bits (SD, XS).
+*
+* Ignore writes to the FS field as set_fpcontext() will set
+* it explicitly.
*/
- if ((mcp->mc_gpregs.gp_sstatus & SSTATUS_SPP) != 0)
+ if (((mcp->mc_gpregs.gp_sstatus ^ tf->tf_sstatus) &
+ ~(SSTATUS_SD | SSTATUS_XS_MASK | SSTATUS_FS_MASK | SSTATUS_UPIE |
+ SSTATUS_UIE)) != 0)
return (EINVAL);
memcpy(tf->tf_t, mcp->mc_gpregs.gp_t, sizeof(tf->tf_t));
@@ -426,7 +431,12 @@ set_fpcontext(struct thread *td, mcontext_t *mcp)
{
#ifdef FPE
struct pcb *curpcb;
+#endif
+ td->td_frame->tf_sstatus &= ~SSTATUS_FS_MASK;
+ td->td_frame->tf_sstatus |= SSTATUS_FS_OFF;
+
+#ifdef FPE
critical_enter();
if ((mcp->mc_flags & _MC_FP_VALID) != 0) {
@@ -436,6 +446,7 @@ set_fpcontext(struct thread *td, mcontext_t *mcp)
sizeof(mcp->mc_fpregs));
curpcb->pcb_fcsr = mcp->mc_fpregs.fp_fcsr;
curpcb->pcb_fpflags = mcp->mc_fpregs.fp_flags & PCB_FP_USERMASK;
+ td->td_frame->tf_sstatus |= SSTATUS_FS_CLEAN;
}
critical_exit();
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r357337 - head/sys/riscv/include
Author: jhb Date: Fri Jan 31 17:49:15 2020 New Revision: 357337 URL: https://svnweb.freebsd.org/changeset/base/357337 Log: Fix 64-bit value of SSTATUS_SD to use an unsigned long. While here, fix MSTATUS_SD to match SSTATUS_SD. Reviewed by: mhorne MFC after:2 weeks Sponsored by: DARPA Differential Revision:https://reviews.freebsd.org/D23434 Modified: head/sys/riscv/include/riscvreg.h Modified: head/sys/riscv/include/riscvreg.h == --- head/sys/riscv/include/riscvreg.h Fri Jan 31 17:40:41 2020 (r357336) +++ head/sys/riscv/include/riscvreg.h Fri Jan 31 17:49:15 2020 (r357337) @@ -73,7 +73,7 @@ #defineSSTATUS_XS_MASK (0x3 << SSTATUS_XS_SHIFT) #defineSSTATUS_SUM (1 << 18) #if __riscv_xlen == 64 -#defineSSTATUS_SD (1 << 63) +#defineSSTATUS_SD (1ul << 63) #else #defineSSTATUS_SD (1 << 31) #endif @@ -110,8 +110,11 @@ #define MSTATUS_VM_SV4810 #define MSTATUS_VM_SV5711 #define MSTATUS_VM_SV6412 -#defineMSTATUS32_SD(1 << 63) -#defineMSTATUS64_SD(1 << 31) +#if __riscv_xlen == 64 +#defineMSTATUS_SD (1ul << 63) +#else +#defineMSTATUS_SD (1 << 31) +#endif #defineMSTATUS_PRV_U 0 /* user */ #defineMSTATUS_PRV_S 1 /* supervisor */ ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r357313 - head/sys/riscv/riscv
Author: jhb Date: Thu Jan 30 22:19:48 2020 New Revision: 357313 URL: https://svnweb.freebsd.org/changeset/base/357313 Log: Trim duplicate CSR swaps from user exceptions. The stack pointer is swapped with the sscratch CSR just before the jump to cpu_exception_handler_user where the first instruction swaps it again. The two swaps together are a no-op, but the csr swap instructions can be expensive (e.g. on Bluespec RISC-V cores csr swap instructions force a full pipeline stall). Reported by: jrtc27 Reviewed by: br MFC after:2 weeks Sponsored by: DARPA Differential Revision:https://reviews.freebsd.org/D23394 Modified: head/sys/riscv/riscv/exception.S Modified: head/sys/riscv/riscv/exception.S == --- head/sys/riscv/riscv/exception.SThu Jan 30 20:05:05 2020 (r357312) +++ head/sys/riscv/riscv/exception.SThu Jan 30 22:19:48 2020 (r357313) @@ -208,7 +208,6 @@ ENTRY(cpu_exception_handler) csrrw sp, sscratch, sp beqzsp, 1f /* User mode detected */ - csrrw sp, sscratch, sp j cpu_exception_handler_user 1: /* Supervisor mode detected */ @@ -225,7 +224,6 @@ ENTRY(cpu_exception_handler_supervisor) END(cpu_exception_handler_supervisor) ENTRY(cpu_exception_handler_user) - csrrw sp, sscratch, sp save_registers 0 mv a0, sp call_C_LABEL(do_trap_user) ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r357305 - in head/sys/riscv: include riscv
Author: jhb
Date: Thu Jan 30 19:15:27 2020
New Revision: 357305
URL: https://svnweb.freebsd.org/changeset/base/357305
Log:
Remove unused fields from struct pcb.
cpu_switch/throw() and savectx() do not save or restore any values in
these fields which mostly held non-callee-save registers.
makectx() copied these fields from kdb_frame, but they weren't used
except for PC_REGS using pcb_sepc. Change PC_REGS to use
kdb_frame->tf_sepc directly instead.
Reviewed by: br
MFC after:2 weeks
Sponsored by: DARPA
Differential Revision:https://reviews.freebsd.org/D23395
Modified:
head/sys/riscv/include/db_machdep.h
head/sys/riscv/include/pcb.h
head/sys/riscv/riscv/genassym.c
head/sys/riscv/riscv/machdep.c
Modified: head/sys/riscv/include/db_machdep.h
==
--- head/sys/riscv/include/db_machdep.h Thu Jan 30 18:49:19 2020
(r357304)
+++ head/sys/riscv/include/db_machdep.h Thu Jan 30 19:15:27 2020
(r357305)
@@ -47,7 +47,7 @@
typedef vm_offset_tdb_addr_t;
typedef long db_expr_t;
-#definePC_REGS() ((db_addr_t)kdb_thrctx->pcb_sepc)
+#definePC_REGS() ((db_addr_t)kdb_frame->tf_sepc)
#defineBKPT_INST (0x00100073)
#defineBKPT_SIZE (INSN_SIZE)
Modified: head/sys/riscv/include/pcb.h
==
--- head/sys/riscv/include/pcb.hThu Jan 30 18:49:19 2020
(r357304)
+++ head/sys/riscv/include/pcb.hThu Jan 30 19:15:27 2020
(r357305)
@@ -46,15 +46,12 @@ struct pcb {
uint64_tpcb_sp; /* Stack pointer */
uint64_tpcb_gp; /* Global pointer */
uint64_tpcb_tp; /* Thread pointer */
- uint64_tpcb_t[7]; /* Temporary registers */
uint64_tpcb_s[12]; /* Saved registers */
- uint64_tpcb_a[8]; /* Argument registers */
uint64_tpcb_x[32][2]; /* Floating point registers */
uint64_tpcb_fcsr; /* Floating point control reg */
uint64_tpcb_fpflags;/* Floating point flags */
#definePCB_FP_STARTED 0x1
#definePCB_FP_USERMASK 0x1
- uint64_tpcb_sepc; /* Supervisor exception pc */
vm_offset_t pcb_onfault;/* Copyinout fault handler */
};
Modified: head/sys/riscv/riscv/genassym.c
==
--- head/sys/riscv/riscv/genassym.c Thu Jan 30 18:49:19 2020
(r357304)
+++ head/sys/riscv/riscv/genassym.c Thu Jan 30 19:15:27 2020
(r357305)
@@ -69,9 +69,7 @@ ASSYM(PCB_RA, offsetof(struct pcb, pcb_ra));
ASSYM(PCB_SP, offsetof(struct pcb, pcb_sp));
ASSYM(PCB_GP, offsetof(struct pcb, pcb_gp));
ASSYM(PCB_TP, offsetof(struct pcb, pcb_tp));
-ASSYM(PCB_T, offsetof(struct pcb, pcb_t));
ASSYM(PCB_S, offsetof(struct pcb, pcb_s));
-ASSYM(PCB_A, offsetof(struct pcb, pcb_a));
ASSYM(PCB_X, offsetof(struct pcb, pcb_x));
ASSYM(PCB_FCSR, offsetof(struct pcb, pcb_fcsr));
Modified: head/sys/riscv/riscv/machdep.c
==
--- head/sys/riscv/riscv/machdep.c Thu Jan 30 18:49:19 2020
(r357304)
+++ head/sys/riscv/riscv/machdep.c Thu Jan 30 19:15:27 2020
(r357305)
@@ -558,15 +558,12 @@ void
makectx(struct trapframe *tf, struct pcb *pcb)
{
- memcpy(pcb->pcb_t, tf->tf_t, sizeof(tf->tf_t));
memcpy(pcb->pcb_s, tf->tf_s, sizeof(tf->tf_s));
- memcpy(pcb->pcb_a, tf->tf_a, sizeof(tf->tf_a));
pcb->pcb_ra = tf->tf_ra;
pcb->pcb_sp = tf->tf_sp;
pcb->pcb_gp = tf->tf_gp;
pcb->pcb_tp = tf->tf_tp;
- pcb->pcb_sepc = tf->tf_sepc;
}
void
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r357302 - head/sys/kern
Author: jhb
Date: Thu Jan 30 18:28:02 2020
New Revision: 357302
URL: https://svnweb.freebsd.org/changeset/base/357302
Log:
Fix use of an uninitialized variable.
ctx (and thus ctx.flags) is stack garbage at the start of this
function, so initialize ctx.flags to an explicit value instead of
using binary operations on the garbage.
Reported by: gcc9
Reviewed by: imp
Differential Revision:https://reviews.freebsd.org/D23368
Modified:
head/sys/kern/subr_stats.c
Modified: head/sys/kern/subr_stats.c
==
--- head/sys/kern/subr_stats.c Thu Jan 30 18:12:41 2020(r357301)
+++ head/sys/kern/subr_stats.c Thu Jan 30 18:28:02 2020(r357302)
@@ -1583,9 +1583,7 @@ stats_v1_blob_iter(struct statsblobv1 *sb, stats_v1_bl
int i, j, firstvoi;
ctx.usrctx = usrctx;
- ctx.flags |= SB_IT_FIRST_CB;
- ctx.flags &= ~(SB_IT_FIRST_VOI | SB_IT_LAST_VOI | SB_IT_FIRST_VOISTAT |
- SB_IT_LAST_VOISTAT);
+ ctx.flags = SB_IT_FIRST_CB;
firstvoi = 1;
for (i = 0; i < NVOIS(sb); i++) {
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r357147 - head/sys/kern
On 1/26/20 6:23 AM, John Baldwin wrote: > Author: jhb > Date: Sun Jan 26 14:23:27 2020 > New Revision: 357147 > URL: https://svnweb.freebsd.org/changeset/base/357147 > > Log: > Revert accidental change from r357146. Oops, this was a different warning fix which I've opened a separate review for (this one was found by gcc 9 instead of clang 10). -- John Baldwin ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r357147 - head/sys/kern
Author: jhb
Date: Sun Jan 26 14:23:27 2020
New Revision: 357147
URL: https://svnweb.freebsd.org/changeset/base/357147
Log:
Revert accidental change from r357146.
Modified:
head/sys/kern/subr_stats.c
Modified: head/sys/kern/subr_stats.c
==
--- head/sys/kern/subr_stats.c Sun Jan 26 14:20:57 2020(r357146)
+++ head/sys/kern/subr_stats.c Sun Jan 26 14:23:27 2020(r357147)
@@ -1583,7 +1583,9 @@ stats_v1_blob_iter(struct statsblobv1 *sb, stats_v1_bl
int i, j, firstvoi;
ctx.usrctx = usrctx;
- ctx.flags = SB_IT_FIRST_CB;
+ ctx.flags |= SB_IT_FIRST_CB;
+ ctx.flags &= ~(SB_IT_FIRST_VOI | SB_IT_LAST_VOI | SB_IT_FIRST_VOISTAT |
+ SB_IT_LAST_VOISTAT);
firstvoi = 1;
for (i = 0; i < NVOIS(sb); i++) {
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r357146 - in head/sys: dev/mrsas dev/sound/pci kern
Author: jhb
Date: Sun Jan 26 14:20:57 2020
New Revision: 357146
URL: https://svnweb.freebsd.org/changeset/base/357146
Log:
Fix some misleading indentation warnings reported by recent clang.
These should not be any functional change. While the change in
emul10kx-pcm.c looks like a real bug fix (as opposed to inconsistent
whitespace), the extra statements were not harmful.
Reviewed by: kib
Sponsored by: DARPA
Differential Revision:https://reviews.freebsd.org/D23363
Modified:
head/sys/dev/mrsas/mrsas_cam.c
head/sys/dev/sound/pci/emu10k1.c
head/sys/dev/sound/pci/emu10kx-pcm.c
head/sys/kern/subr_stats.c
Modified: head/sys/dev/mrsas/mrsas_cam.c
==
--- head/sys/dev/mrsas/mrsas_cam.c Sun Jan 26 14:19:08 2020
(r357145)
+++ head/sys/dev/mrsas/mrsas_cam.c Sun Jan 26 14:20:57 2020
(r357146)
@@ -1908,13 +1908,14 @@ mrsas_track_scsiio(struct mrsas_softc *sc, target_id_t
for (i = 0 ; i < sc->max_fw_cmds; i++) {
mpt_cmd = sc->mpt_cmd_list[i];
- /*
-* Check if the target_id and bus_id is same as the timeout IO
-*/
- if (mpt_cmd->ccb_ptr) {
- /* bus_id = 1 denotes a VD */
- if (bus_id == 1)
- tgt_id = (mpt_cmd->ccb_ptr->ccb_h.target_id -
(MRSAS_MAX_PD - 1));
+ /*
+* Check if the target_id and bus_id is same as the timeout IO
+*/
+ if (mpt_cmd->ccb_ptr) {
+ /* bus_id = 1 denotes a VD */
+ if (bus_id == 1)
+ tgt_id =
+ (mpt_cmd->ccb_ptr->ccb_h.target_id -
(MRSAS_MAX_PD - 1));
if (mpt_cmd->ccb_ptr->cpi.bus_id == bus_id &&
mpt_cmd->ccb_ptr->ccb_h.target_id == tgt_id) {
Modified: head/sys/dev/sound/pci/emu10k1.c
==
--- head/sys/dev/sound/pci/emu10k1.cSun Jan 26 14:19:08 2020
(r357145)
+++ head/sys/dev/sound/pci/emu10k1.cSun Jan 26 14:20:57 2020
(r357146)
@@ -1257,11 +1257,12 @@ emu_intr(void *data)
#endif
}
- if (stat & EMU_IPR_MIDIRECVBUFE)
- if (sc->mpu_intr) {
- (sc->mpu_intr)(sc->mpu);
- ack |= EMU_IPR_MIDIRECVBUFE | EMU_IPR_MIDITRANSBUFE;
- }
+ if (stat & EMU_IPR_MIDIRECVBUFE) {
+ if (sc->mpu_intr) {
+ (sc->mpu_intr)(sc->mpu);
+ ack |= EMU_IPR_MIDIRECVBUFE |
EMU_IPR_MIDITRANSBUFE;
+ }
+ }
if (stat & ~ack)
device_printf(sc->dev, "dodgy irq: %x (harmless)\n",
stat & ~ack);
Modified: head/sys/dev/sound/pci/emu10kx-pcm.c
==
--- head/sys/dev/sound/pci/emu10kx-pcm.cSun Jan 26 14:19:08 2020
(r357145)
+++ head/sys/dev/sound/pci/emu10kx-pcm.cSun Jan 26 14:20:57 2020
(r357146)
@@ -263,11 +263,12 @@ emu_dspmixer_uninit(struct snd_mixer *m)
/* drop submixer for AC97 codec */
sc = mix_getdevinfo(m);
- if (sc->sm != NULL)
+ if (sc->sm != NULL) {
err = mixer_delete(sc->sm);
if (err)
return (err);
sc->sm = NULL;
+ }
return (0);
}
Modified: head/sys/kern/subr_stats.c
==
--- head/sys/kern/subr_stats.c Sun Jan 26 14:19:08 2020(r357145)
+++ head/sys/kern/subr_stats.c Sun Jan 26 14:20:57 2020(r357146)
@@ -1583,9 +1583,7 @@ stats_v1_blob_iter(struct statsblobv1 *sb, stats_v1_bl
int i, j, firstvoi;
ctx.usrctx = usrctx;
- ctx.flags |= SB_IT_FIRST_CB;
- ctx.flags &= ~(SB_IT_FIRST_VOI | SB_IT_LAST_VOI | SB_IT_FIRST_VOISTAT |
- SB_IT_LAST_VOISTAT);
+ ctx.flags = SB_IT_FIRST_CB;
firstvoi = 1;
for (i = 0; i < NVOIS(sb); i++) {
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r357145 - head/sys/conf
Author: jhb
Date: Sun Jan 26 14:19:08 2020
New Revision: 357145
URL: https://svnweb.freebsd.org/changeset/base/357145
Log:
Compile hack.c with normal CFLAGS + -shared -nostdlib.
Originally, hack.c was compiled into a shard object with just -shared
-nostdlib. This assumed that ${CC} did not require any additional
flags for ABIs, cross-building, etc.
When kern.post.mk was created in r89509 by reducing duplication in
kernel Makefile. files, the -shared flag was moved into a
HACK_EXTRA_FLAGS variable so that sparc64 could override it with
-Wl,-shared. The sparc64 hack was removed in r111650, but
HACK_EXTRA_FLAGS was left in place. Over time, we have started
support toolchains that require flags to support alternate ABIs on
MIPS and PowerPC and started (ab)using HACK_EXTRA_FLAGS to set only
those flags.
I need to fix risc-v to pass -mno-relax to the hack.c build for lld in
llvm 10, and the patches to support cross-build from non-FreeBSD hosts
need to include -target for clang in CFLAGS for hack.c. Rather than
adding more hacks into HACK_EXTRA_FLAGS, just use the full set of
CFLAGS with hack.c.
Reviewed by: kib, arichardson
MFC after:1 month
Sponsored by: DARPA
Differential Revision:https://reviews.freebsd.org/D23362
Modified:
head/sys/conf/Makefile.mips
head/sys/conf/Makefile.powerpc
head/sys/conf/kern.post.mk
Modified: head/sys/conf/Makefile.mips
==
--- head/sys/conf/Makefile.mips Sun Jan 26 11:54:21 2020(r357144)
+++ head/sys/conf/Makefile.mips Sun Jan 26 14:19:08 2020(r357145)
@@ -50,12 +50,9 @@ ARCH_FLAGS+=-mabi=${MIPS_ABI}
EXTRA_FLAGS=-fno-pic -mno-abicalls -G0 -DKERNLOADADDR=${KERNLOADADDR}
EXTRA_FLAGS+=-${MIPS_ENDIAN}
-HACK_EXTRA_FLAGS=-shared
-
# We add the -fno-pic flag to kernels because otherwise performance
# is extremely poor, as well as -mno-abicalls to force no ABI usage.
CFLAGS+=${EXTRA_FLAGS} $(ARCH_FLAGS)
-HACK_EXTRA_FLAGS+=${EXTRA_FLAGS} $(ARCH_FLAGS)
TRAMP_ARCH_FLAGS?=$(ARCH_FLAGS)
TRAMP_EXTRA_FLAGS=${EXTRA_FLAGS} ${TRAMP_ARCH_FLAGS}
# Kernel code is always compiled with soft-float on MIPS
Modified: head/sys/conf/Makefile.powerpc
==
--- head/sys/conf/Makefile.powerpc Sun Jan 26 11:54:21 2020
(r357144)
+++ head/sys/conf/Makefile.powerpc Sun Jan 26 14:19:08 2020
(r357145)
@@ -39,7 +39,6 @@ INCLUDES+= -I$S/contrib/libfdt
# Force __SPE__, since the builtin will be removed later with -mno-spe
CFLAGS.gcc+= -mabi=spe -D__SPE__
CFLAGS.clang+= -mspe -D__SPE__ -m32
-HACK_EXTRA_FLAGS= -shared -m32 -mspe -D__SPE__
.endif
CFLAGS+= -msoft-float
CFLAGS.gcc+= -Wa,-many
Modified: head/sys/conf/kern.post.mk
==
--- head/sys/conf/kern.post.mk Sun Jan 26 11:54:21 2020(r357144)
+++ head/sys/conf/kern.post.mk Sun Jan 26 14:19:08 2020(r357145)
@@ -226,10 +226,9 @@ kernel-clean:
# This is a hack. BFD "optimizes" away dynamic mode if there are no
# dynamic references. We could probably do a '-Bforcedynamic' mode like
# in the a.out ld. For now, this works.
-HACK_EXTRA_FLAGS?= -shared
hack.pico: Makefile
:> hack.c
- ${CC} ${HACK_EXTRA_FLAGS} -nostdlib hack.c -o hack.pico
+ ${CC} -shared ${CFLAGS} -nostdlib hack.c -o hack.pico
rm -f hack.c
offset.inc: $S/kern/genoffset.sh genoffset.o
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r357051 - head/sys/dev/bge
On 1/23/20 7:32 PM, Gleb Smirnoff wrote: > On Thu, Jan 23, 2020 at 05:09:14PM -1000, Jeff Roberson wrote: > J> While we don't have a policy strictly requiring reviews it is the norm to > J> have substantial changes socialized and reviewed. I appreciate the work > J> that you are doing but it likely should've been discussed somewhere > J> more publicly. I apologized if I missed it but I don't see reference to > J> anything. > > That was https://reviews.freebsd.org/D23242 A review alone isn't sufficient for large, sweeping changes in my mind. For major changes, a thread on arch@ or net@ or the like is probably more appropriate. You can include a link to a review or git branch, etc. in that e-mail, but phabricator aren't as well suited to higher-level design-review type discussion, more for implementation-review. > J> Architecturally I am more concerned with the coarseness of net_epoch and > J> the duration of hold becoming a resource utilization problem in high > J> turn-over workloads. Like short connection tcp. Has anyone done > J> substantial testing here? epoch as it is today will hold every free > J> callback for a minimum of several clock ticks and a maximum of 2x the > J> duration of the longest epoch section time. With preemption, etc. this > J> could be 100s of ms of PCBs held. > > We also are concerned about that theoretically. Haven't yet seen effect > in practice, but our sessions are mostly longer living. First we have the > tunable to limit batching. Second, there are some ideas on how to improve > the garbage collector performance if it becomes an issue. There are other workloads than Netflix. ;) Verisign has incredibly short-lived connections with very high turnover. I think though that they have already abandoned the in-tree network stack for a userland stack built on netmap. Still, I think that there are probably other FreeBSD users that are probably somewhere in the middle that shouldn't be ignored. Packet batching would not be impossible by simply using m_nextpkt chains in mbufs passed up to ether_input and having ether_input pass them in a loop to the next higher loop (as a first step). That would reduce unlock/lock operations in drivers (for those still using locks on receive) as well as permitting ether_input to process batches under a single epoch invocation. -- John Baldwin ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r357051 - head/sys/dev/bge
On 1/23/20 9:15 PM, Gleb Smirnoff wrote: > On Thu, Jan 23, 2020 at 06:18:15PM -1000, Jeff Roberson wrote: > J> > That was https://reviews.freebsd.org/D23242 > J> > J> Ok thank you. Can you tag commits so people can see the discussion? Was > J> it in one I missed? When I'm committing a long patch series I include the > J> link in all of them. > > I probably now on will also include in every patch in a serie. I just > dislike that first one (usually a preparation change) closes the review. You can also add additional commits to a review via 'Edit related revisions' in the web ui if you forget to tag a commit. -- John Baldwin ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r357062 - head/share/man/man9
Author: jhb Date: Thu Jan 23 23:36:58 2020 New Revision: 357062 URL: https://svnweb.freebsd.org/changeset/base/357062 Log: Correct the return types of fueword*(). MFC after:1 week Sponsored by: DARPA Modified: head/share/man/man9/fetch.9 Modified: head/share/man/man9/fetch.9 == --- head/share/man/man9/fetch.9 Thu Jan 23 22:13:41 2020(r357061) +++ head/share/man/man9/fetch.9 Thu Jan 23 23:36:58 2020(r357062) @@ -34,7 +34,7 @@ .\" .\" $FreeBSD$ .\" -.Dd April 17, 2018 +.Dd January 23, 2020 .Dt FETCH 9 .Os .Sh NAME @@ -61,11 +61,11 @@ .Fn fuword32 "volatile const void *base" .Ft int64_t .Fn fuword64 "volatile const void *base" -.Ft long +.Ft int .Fn fueword "volatile const void *base" "long *val" -.Ft int32_t +.Ft int .Fn fueword32 "volatile const void *base" "int32_t *val" -.Ft int64_t +.Ft int .Fn fueword64 "volatile const void *base" "int64_t *val" .In sys/resourcevar.h .Sh DESCRIPTION ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r356996 - head
Author: jhb
Date: Wed Jan 22 21:21:24 2020
New Revision: 356996
URL: https://svnweb.freebsd.org/changeset/base/356996
Log:
Remove support for auto-selecting an external binutils.
All of the in-tree architectures not supported by in-tree binutils are
supported by lld, so the condition is now always false. It also
didn't fully work since the external binutils are installed into a
directory that uses the host's OS version, not the target OS version.
Reviewed by: emaste, imp
Differential Revision:https://reviews.freebsd.org/D23294
Modified:
head/Makefile.inc1
Modified: head/Makefile.inc1
==
--- head/Makefile.inc1 Wed Jan 22 21:01:19 2020(r356995)
+++ head/Makefile.inc1 Wed Jan 22 21:21:24 2020(r356996)
@@ -180,24 +180,6 @@ MK_SYSTEM_LINKER= no
.if defined(CROSS_TOOLCHAIN_PREFIX)
CROSS_BINUTILS_PREFIX?=${CROSS_TOOLCHAIN_PREFIX}
.endif
-# If we do not have a bootstrap binutils (because the in-tree one does not
-# support the target architecture), provide a default cross-binutils prefix.
-# This allows riscv64 builds, for example, to automatically use the
-# riscv64-binutils port or package.
-.if !make(showconfig) && !defined(_NO_INCLUDE_COMPILERMK)
-.if !empty(BROKEN_OPTIONS:MBINUTILS_BOOTSTRAP) && \
-${MK_LLD_BOOTSTRAP} == "no" && \
-!defined(CROSS_BINUTILS_PREFIX)
-CROSS_BINUTILS_PREFIX=/usr/local/${TARGET_TRIPLE}/bin/
-.if !exists(${CROSS_BINUTILS_PREFIX})
-.if !empty(BROKEN_OPTIONS:MGCC_BOOTSTRAP) && ${MK_CLANG_BOOTSTRAP} == "no"
-.error In-tree toolchain does not support the ${TARGET_ARCH} architecture.
Install the ${TARGET_ARCH}-xtoolchain-gcc port or package or set
CROSS_TOOLCHAIN_PREFIX.
-.else
-.error In-tree binutils does not support the ${TARGET_ARCH} architecture.
Install the ${TARGET_ARCH}-binutils port or package or set
CROSS_BINUTILS_PREFIX.
-.endif
-.endif
-.endif
-.endif
XBINUTILS= AS AR LD NM OBJCOPY RANLIB SIZE STRINGS
.for BINUTIL in ${XBINUTILS}
.if defined(CROSS_BINUTILS_PREFIX) && \
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r356911 - stable/12/sys/opencrypto
Author: jhb
Date: Mon Jan 20 11:54:00 2020
New Revision: 356911
URL: https://svnweb.freebsd.org/changeset/base/356911
Log:
MFC 356561: Add stricter checking on mac key lengths.
Negative lengths are always invalid. The key length should also
be zero for hash algorithms that do not accept a key.
admbugs: 949
Modified:
stable/12/sys/opencrypto/cryptodev.c
Directory Properties:
stable/12/ (props changed)
Modified: stable/12/sys/opencrypto/cryptodev.c
==
--- stable/12/sys/opencrypto/cryptodev.cMon Jan 20 11:45:18 2020
(r356910)
+++ stable/12/sys/opencrypto/cryptodev.cMon Jan 20 11:54:00 2020
(r356911)
@@ -585,8 +585,8 @@ cryptof_ioctl(
if (thash) {
cria.cri_alg = thash->type;
cria.cri_klen = sop->mackeylen * 8;
- if (thash->keysize != 0 &&
- sop->mackeylen > thash->keysize) {
+ if (sop->mackeylen > thash->keysize ||
+ sop->mackeylen < 0) {
CRYPTDEB("invalid mac key length");
error = EINVAL;
SDT_PROBE1(opencrypto, dev, ioctl, error,
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r356908 - in stable: 11/sys/opencrypto 12/sys/opencrypto
Author: jhb
Date: Mon Jan 20 11:19:55 2020
New Revision: 356908
URL: https://svnweb.freebsd.org/changeset/base/356908
Log:
MFC 356507,356520: Add a reference count to cryptodev sessions.
356507:
Add a reference count to cryptodev sessions.
This prevents use-after-free races with crypto requests (which may
sleep) and CIOCFSESSION as well as races from current CIOCFSESSION
requests.
356520:
Remove no-longer-used function prototype.
admbugs: 949
Sponsored by: Chelsio Communications
Modified:
stable/12/sys/opencrypto/cryptodev.c
Directory Properties:
stable/12/ (props changed)
Changes in other areas also in this revision:
Modified:
stable/11/sys/opencrypto/cryptodev.c
Directory Properties:
stable/11/ (props changed)
Modified: stable/12/sys/opencrypto/cryptodev.c
==
--- stable/12/sys/opencrypto/cryptodev.cMon Jan 20 09:16:06 2020
(r356907)
+++ stable/12/sys/opencrypto/cryptodev.cMon Jan 20 11:19:55 2020
(r356908)
@@ -266,6 +266,7 @@ crypt_kop_to_32(const struct crypt_kop *from, struct c
struct csession {
TAILQ_ENTRY(csession) next;
crypto_session_t cses;
+ volatile u_int refs;
u_int32_t ses;
struct mtx lock; /* for op submission */
@@ -292,6 +293,7 @@ struct cryptop_data {
struct fcrypt {
TAILQ_HEAD(csessionlist, csession) csessions;
int sesn;
+ struct mtx lock;
};
static struct timeval warninterval = { .tv_sec = 60, .tv_usec = 0 };
@@ -323,8 +325,7 @@ static struct fileops cryptofops = {
};
static struct csession *csefind(struct fcrypt *, u_int);
-static int csedelete(struct fcrypt *, struct csession *);
-static struct csession *cseadd(struct fcrypt *, struct csession *);
+static bool csedelete(struct fcrypt *, u_int);
static struct csession *csecreate(struct fcrypt *, crypto_session_t, caddr_t,
u_int64_t, caddr_t, u_int64_t, u_int32_t, u_int32_t, struct enc_xform *,
struct auth_hash *);
@@ -668,13 +669,10 @@ bail:
break;
case CIOCFSESSION:
ses = *(u_int32_t *)data;
- cse = csefind(fcr, ses);
- if (cse == NULL) {
+ if (!csedelete(fcr, ses)) {
SDT_PROBE1(opencrypto, dev, ioctl, error, __LINE__);
return (EINVAL);
}
- csedelete(fcr, cse);
- csefree(cse);
break;
case CIOCCRYPT:
#ifdef COMPAT_FREEBSD32
@@ -691,6 +689,7 @@ bail:
return (EINVAL);
}
error = cryptodev_op(cse, cop, active_cred, td);
+ csefree(cse);
#ifdef COMPAT_FREEBSD32
if (error == 0 && cmd == CIOCCRYPT32)
crypt_op_to_32(cop, data);
@@ -757,6 +756,7 @@ bail:
return (EINVAL);
}
error = cryptodev_aead(cse, caead, active_cred, td);
+ csefree(cse);
break;
default:
error = EINVAL;
@@ -1375,6 +1375,9 @@ cryptof_close(struct file *fp, struct thread *td)
while ((cse = TAILQ_FIRST(>csessions))) {
TAILQ_REMOVE(>csessions, cse, next);
+ KASSERT(cse->refs == 1,
+ ("%s: crypto session %p with %d refs", __func__, cse,
+ cse->refs));
csefree(cse);
}
free(fcr, M_XDATA);
@@ -1395,34 +1398,36 @@ csefind(struct fcrypt *fcr, u_int ses)
{
struct csession *cse;
- TAILQ_FOREACH(cse, >csessions, next)
- if (cse->ses == ses)
+ mtx_lock(>lock);
+ TAILQ_FOREACH(cse, >csessions, next) {
+ if (cse->ses == ses) {
+ refcount_acquire(>refs);
+ mtx_unlock(>lock);
return (cse);
+ }
+ }
+ mtx_unlock(>lock);
return (NULL);
}
-static int
-csedelete(struct fcrypt *fcr, struct csession *cse_del)
+static bool
+csedelete(struct fcrypt *fcr, u_int ses)
{
struct csession *cse;
+ mtx_lock(>lock);
TAILQ_FOREACH(cse, >csessions, next) {
- if (cse == cse_del) {
+ if (cse->ses == ses) {
TAILQ_REMOVE(>csessions, cse, next);
- return (1);
+ mtx_unlock(>lock);
+ csefree(cse);
+ return (true);
}
}
- return (0);
+ mtx_unlock(>lock);
+ return (false);
}
-static struct csession *
-cseadd(struct fcrypt *fcr, struct csession *cse)
-{
- TAILQ_INSERT_TAIL(>csessions, cse, next);
- cse->ses = fcr->sesn++;
- return (cse);
-}
-
struct csession *
csecreate(struct fcrypt *fcr, crypto_session_t cses, caddr_t key, u_int64_t
svn commit: r356908 - in stable: 11/sys/opencrypto 12/sys/opencrypto
Author: jhb
Date: Mon Jan 20 11:19:55 2020
New Revision: 356908
URL: https://svnweb.freebsd.org/changeset/base/356908
Log:
MFC 356507,356520: Add a reference count to cryptodev sessions.
356507:
Add a reference count to cryptodev sessions.
This prevents use-after-free races with crypto requests (which may
sleep) and CIOCFSESSION as well as races from current CIOCFSESSION
requests.
356520:
Remove no-longer-used function prototype.
admbugs: 949
Sponsored by: Chelsio Communications
Modified:
stable/11/sys/opencrypto/cryptodev.c
Directory Properties:
stable/11/ (props changed)
Changes in other areas also in this revision:
Modified:
stable/12/sys/opencrypto/cryptodev.c
Directory Properties:
stable/12/ (props changed)
Modified: stable/11/sys/opencrypto/cryptodev.c
==
--- stable/11/sys/opencrypto/cryptodev.cMon Jan 20 09:16:06 2020
(r356907)
+++ stable/11/sys/opencrypto/cryptodev.cMon Jan 20 11:19:55 2020
(r356908)
@@ -268,6 +268,7 @@ crypt_kop_to_32(const struct crypt_kop *from, struct c
struct csession {
TAILQ_ENTRY(csession) next;
u_int64_t sid;
+ volatile u_int refs;
u_int32_t ses;
struct mtx lock; /* for op submission */
@@ -294,6 +295,7 @@ struct cryptop_data {
struct fcrypt {
TAILQ_HEAD(csessionlist, csession) csessions;
int sesn;
+ struct mtx lock;
};
static struct timeval warninterval = { .tv_sec = 60, .tv_usec = 0 };
@@ -325,8 +327,7 @@ static struct fileops cryptofops = {
};
static struct csession *csefind(struct fcrypt *, u_int);
-static int csedelete(struct fcrypt *, struct csession *);
-static struct csession *cseadd(struct fcrypt *, struct csession *);
+static int csedelete(struct fcrypt *, u_int);
static struct csession *csecreate(struct fcrypt *, u_int64_t, caddr_t,
u_int64_t, caddr_t, u_int64_t, u_int32_t, u_int32_t, struct enc_xform *,
struct auth_hash *);
@@ -617,13 +618,9 @@ bail:
break;
case CIOCFSESSION:
ses = *(u_int32_t *)data;
- cse = csefind(fcr, ses);
- if (cse == NULL) {
+ error = csedelete(fcr, ses);
+ if (error != 0)
SDT_PROBE1(opencrypto, dev, ioctl, error, __LINE__);
- return (EINVAL);
- }
- csedelete(fcr, cse);
- error = csefree(cse);
break;
case CIOCCRYPT:
#ifdef COMPAT_FREEBSD32
@@ -640,6 +637,7 @@ bail:
return (EINVAL);
}
error = cryptodev_op(cse, cop, active_cred, td);
+ (void)csefree(cse);
#ifdef COMPAT_FREEBSD32
if (error == 0 && cmd == CIOCCRYPT32)
crypt_op_to_32(cop, data);
@@ -706,6 +704,7 @@ bail:
return (EINVAL);
}
error = cryptodev_aead(cse, caead, active_cred, td);
+ (void)csefree(cse);
break;
default:
error = EINVAL;
@@ -1323,6 +1322,9 @@ cryptof_close(struct file *fp, struct thread *td)
while ((cse = TAILQ_FIRST(>csessions))) {
TAILQ_REMOVE(>csessions, cse, next);
+ KASSERT(cse->refs == 1,
+ ("%s: crypto session %p with %d refs", __func__, cse,
+ cse->refs));
(void)csefree(cse);
}
free(fcr, M_XDATA);
@@ -1343,34 +1345,35 @@ csefind(struct fcrypt *fcr, u_int ses)
{
struct csession *cse;
- TAILQ_FOREACH(cse, >csessions, next)
- if (cse->ses == ses)
+ mtx_lock(>lock);
+ TAILQ_FOREACH(cse, >csessions, next) {
+ if (cse->ses == ses) {
+ refcount_acquire(>refs);
+ mtx_unlock(>lock);
return (cse);
+ }
+ }
+ mtx_unlock(>lock);
return (NULL);
}
static int
-csedelete(struct fcrypt *fcr, struct csession *cse_del)
+csedelete(struct fcrypt *fcr, u_int ses)
{
struct csession *cse;
+ mtx_lock(>lock);
TAILQ_FOREACH(cse, >csessions, next) {
- if (cse == cse_del) {
+ if (cse->ses == ses) {
TAILQ_REMOVE(>csessions, cse, next);
- return (1);
+ mtx_unlock(>lock);
+ return (csefree(cse));
}
}
- return (0);
+ mtx_unlock(>lock);
+ return (EINVAL);
}
-static struct csession *
-cseadd(struct fcrypt *fcr, struct csession *cse)
-{
- TAILQ_INSERT_TAIL(>csessions, cse, next);
- cse->ses = fcr->sesn++;
- return (cse);
-}
-
struct csession *
csecreate(struct fcrypt *fcr, u_int64_t sid, caddr_t key, u_int64_t
svn commit: r356875 - stable/11/share/man/man7
Author: jhb Date: Sat Jan 18 23:46:50 2020 New Revision: 356875 URL: https://svnweb.freebsd.org/changeset/base/356875 Log: MFC 356274: Some minor tweaks to arch(7). - Drop mention of _LP64. FreeBSD's source generally uses __LP64__ instead of _LP64, and the relevant macros are better covered in the "Predefined Macros" section. Modified: stable/11/share/man/man7/arch.7 Directory Properties: stable/11/ (props changed) Modified: stable/11/share/man/man7/arch.7 == --- stable/11/share/man/man7/arch.7 Sat Jan 18 23:46:30 2020 (r356874) +++ stable/11/share/man/man7/arch.7 Sat Jan 18 23:46:50 2020 (r356875) @@ -139,12 +139,6 @@ and are 8 bytes. .El .Pp -Compilers define the -.Dv _LP64 -symbol when compiling for an -.Dv LP64 -ABI. -.Pp Some machines support more that one .Fx ABI. ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r356874 - stable/12/share/man/man7
Author: jhb Date: Sat Jan 18 23:46:30 2020 New Revision: 356874 URL: https://svnweb.freebsd.org/changeset/base/356874 Log: MFC 356274: Some minor tweaks to arch(7). - Drop mention of _LP64. FreeBSD's source generally uses __LP64__ instead of _LP64, and the relevant macros are better covered in the "Predefined Macros" section. - Fix a noun/verb disagreement. Modified: stable/12/share/man/man7/arch.7 Directory Properties: stable/12/ (props changed) Modified: stable/12/share/man/man7/arch.7 == --- stable/12/share/man/man7/arch.7 Sat Jan 18 23:43:36 2020 (r356873) +++ stable/12/share/man/man7/arch.7 Sat Jan 18 23:46:30 2020 (r356874) @@ -140,12 +140,6 @@ and are 8 bytes. .El .Pp -Compilers define the -.Dv _LP64 -symbol when compiling for an -.Dv LP64 -ABI. -.Pp Some machines support more than one .Fx ABI. @@ -401,7 +395,7 @@ These variables are not otherwise documented and are u in the build system. .Bl -tag -width "MACHINE_CPUARCH" .It Dv MACHINE -Represent the hardware platform. +Represents the hardware platform. This is the same as the native platform's .Xr uname 1 .Fl m ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r356872 - stable/12/share/man/man7
Author: jhb Date: Sat Jan 18 23:42:57 2020 New Revision: 356872 URL: https://svnweb.freebsd.org/changeset/base/356872 Log: MFC 356209: Formatting fixes for tables, no content changes. - Add missing .Pp after the end of some lists so that there is a blank line before the subsequent paragraph. - Use a more typical '-tag' bullet list of the make variable descriptions at the end. This adds separation between bullets and is the formatting typically used in manpages for this sort of list. Modified: stable/12/share/man/man7/arch.7 Directory Properties: stable/12/ (props changed) Modified: stable/12/share/man/man7/arch.7 == --- stable/12/share/man/man7/arch.7 Sat Jan 18 23:30:42 2020 (r356871) +++ stable/12/share/man/man7/arch.7 Sat Jan 18 23:42:57 2020 (r356872) @@ -139,6 +139,7 @@ and .Vt void * are 8 bytes. .El +.Pp Compilers define the .Dv _LP64 symbol when compiling for an @@ -162,6 +163,7 @@ Examples are: .It Dv powerpc64 Ta Dv powerpc .It Dv mips64* Ta Dv mips* .El +.Pp .Dv aarch64 currently does not support execution of .Dv armv6 @@ -181,6 +183,7 @@ On all supported architectures: .It float Ta 4 .It double Ta 8 .El +.Pp Integers are represented in two's complement. Alignment of integer and pointer types is natural, that is, the address of the variable must be congruent to zero modulo the type size. @@ -396,8 +399,9 @@ Most of the externally settable variables are defined man page. These variables are not otherwise documented and are used extensively in the build system. -.Bl -column -offset indent "Sy Variable" "Sy Meaning and usage" -.It Dv MACHINE Represent the hardware platform. +.Bl -tag -width "MACHINE_CPUARCH" +.It Dv MACHINE +Represent the hardware platform. This is the same as the native platform's .Xr uname 1 .Fl m @@ -432,7 +436,8 @@ Generally, .Dv MACHINE should only be used in src/sys and src/stand or in system imagers or installers. -.It Dv MACHINE_ARCHRepresents the CPU processor architecture. +.It Dv MACHINE_ARCH +Represents the CPU processor architecture. This is the same as the native platforms .Xr uname 1 .Fl p @@ -463,7 +468,8 @@ It is unfortunate that amd64 specifies the 64-bit evol platform (it matches the 'first rule') as everybody else uses x86_64. There is no standard name for the processor: each OS selects its own conventions. -.It Dv MACHINE_CPUARCH Represents the source location for a given +.It Dv MACHINE_CPUARCH +Represents the source location for a given .Dv MACHINE_ARCH . For example, .Dv MACHINE_CPUARCH @@ -475,25 +481,29 @@ The FreeBSD source base supports amd64 and i386 with t distinct source bases living in subdirectories named amd64 and i386 (though behind the scenes there's some sharing that fits into this framework). -.It Dv CPUTYPE Sets the flavor of +.It Dv CPUTYPE +Sets the flavor of .Dv MACHINE_ARCH to build. It is used to optimize the build for a specific CPU / core that the binaries run on. Generally, this does not change the ABI, though it can be a fine line between optimization for specific cases. -.It Dv TARGET Used to set +.It Dv TARGET +Used to set .Dv MACHINE in the top level Makefile for cross building. Unused outside of that scope. It is not passed down to the rest of the build. Makefiles outside of the top level should not use it at all (though some have their own private copy for hysterical raisons). -.It Dv TARGET_ARCH Used to set +.It Dv TARGET_ARCH +Used to set .Dv MACHINE_ARCH by the top level Makefile for cross building. Like -.Dv TARGET , it is unused outside of that scope. +.Dv TARGET , +it is unused outside of that scope. .El .Sh SEE ALSO .Xr src.conf 5 , ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r356873 - stable/11/share/man/man7
Author: jhb Date: Sat Jan 18 23:43:36 2020 New Revision: 356873 URL: https://svnweb.freebsd.org/changeset/base/356873 Log: MFC 356209: Formatting fixes for tables, no content changes. - Add missing .Pp after the end of some lists so that there is a blank line before the subsequent paragraph. Modified: stable/11/share/man/man7/arch.7 Directory Properties: stable/11/ (props changed) Modified: stable/11/share/man/man7/arch.7 == --- stable/11/share/man/man7/arch.7 Sat Jan 18 23:42:57 2020 (r356872) +++ stable/11/share/man/man7/arch.7 Sat Jan 18 23:43:36 2020 (r356873) @@ -138,6 +138,7 @@ and .Vt void * are 8 bytes. .El +.Pp Compilers define the .Dv _LP64 symbol when compiling for an @@ -161,6 +162,7 @@ Examples are: .It Dv powerpc64 Ta Dv powerpc .It Dv mips64* Ta Dv mips* .El +.Pp .Dv aarch64 currently does not support execution of .Dv armv6 @@ -178,6 +180,7 @@ On all supported architectures: .It float Ta 4 .It double Ta 8 .El +.Pp Integers are represented in two's complement. Alignment of integer and pointer types is natural, that is, the address of the variable must be congruent to zero modulo the type size. ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r356871 - in stable/12/share/man: man4 man9
Author: jhb
Date: Sat Jan 18 23:30:42 2020
New Revision: 356871
URL: https://svnweb.freebsd.org/changeset/base/356871
Log:
MFC 355866,355871: Update the crypto(4) and crypto(9) manpages.
355866:
Update the crypto(4) and crypto(9) manpages.
There are probably bits that are still wrong, but this fixes some
things at least:
- Add named arguments to the functions in crypto(9).
- Add missing algorithms.
- Don't mention arguments that don't exist in crypto_register.
- Add CIOGSESSION2.
- Remove CIOCNFSESSION.
- Clarify some stale language that assumed an fd had only one sesson.
- Note that you have to use CRIOGET and add a note in BUGS lamenting
that one has to use CRIOGET.
- Various other cleanups.
355871:
Bump Dd for changes in r355866.
Sponsored by: Chelsio Communications
Modified:
stable/12/share/man/man4/crypto.4
stable/12/share/man/man9/crypto.9
Directory Properties:
stable/12/ (props changed)
Modified: stable/12/share/man/man4/crypto.4
==
--- stable/12/share/man/man4/crypto.4 Sat Jan 18 23:26:15 2020
(r356870)
+++ stable/12/share/man/man4/crypto.4 Sat Jan 18 23:30:42 2020
(r356871)
@@ -60,7 +60,7 @@
.\"
.\" $FreeBSD$
.\"
-.Dd September 21, 2017
+.Dd December 17, 2019
.Dt CRYPTO 4
.Os
.Sh NAME
@@ -78,7 +78,7 @@
The
.Nm
driver gives user-mode applications access to hardware-accelerated
-cryptographic transforms, as implemented by the
+cryptographic transforms as implemented by the
.Xr crypto 9
in-kernel interface.
.Pp
@@ -87,15 +87,15 @@ The
special device provides an
.Xr ioctl 2
based interface.
-User-mode applications should open the special device,
+User-mode applications open the special device and
then issue
.Xr ioctl 2
calls on the descriptor.
User-mode access to
.Pa /dev/crypto
-is controlled by three
+is controlled by two
.Xr sysctl 8
-variables,
+variables:
.Ic kern.userasymcrypto
and
.Ic kern.cryptodevallowsoft .
@@ -103,8 +103,8 @@ and
The
.Nm
device provides two distinct modes of operation: one mode for
-symmetric-keyed cryptographic requests, and a second mode for
-both asymmetric-key (public-key/private-key) requests, and for
+symmetric-keyed cryptographic requests and digests, and a second mode for
+both asymmetric-key (public-key/private-key) requests and
modular arithmetic (for Diffie-Hellman key exchange and other
cryptographic protocols).
The two modes are described separately below.
@@ -113,12 +113,22 @@ Regardless of whether symmetric-key or asymmetric-key
to be performed, use of the device requires a basic series of steps:
.Bl -enum
.It
-Open a file descriptor for the device.
-See
-.Xr open 2 .
+Open the
+.Pa /dev/crypto
+device.
.It
-If any symmetric operation will be performed,
-create one session, with
+Create a new cryptography file descriptor via
+.Dv CRIOGET
+to use for all subsequent
+.Xr ioctl 2
+commands.
+.It
+Close the
+.Pa /dev/crypto
+device.
+.It
+If any symmetric-keyed cryptographic or digest operations will be performed,
+create a session with
.Dv CIOCGSESSION .
Most applications will require at least one symmetric session.
Since cipher and MAC keys are tied to sessions, many
@@ -134,11 +144,13 @@ or
.Dv CIOCKEY
(asymmetric).
.It
-Destroy one session with
+Optionally destroy a session with
.Dv CIOCFSESSION .
.It
-Close the device with
+Close the cryptography file descriptor with
.Xr close 2 .
+This will automatically close any remaining sessions associated with the
+file desriptor.
.El
.Sh SYMMETRIC-KEY OPERATION
The symmetric-key operation mode provides a context-based API
@@ -196,9 +208,9 @@ struct session_op {
u_int32_t mac; /* e.g. CRYPTO_MD5_HMAC */
u_int32_t keylen; /* cipher key */
-void * key;
+const void *key;
int mackeylen; /* mac key */
-void * mackey;
+const void *mackey;
u_int32_t ses; /* returns: ses # */
};
@@ -241,12 +253,36 @@ and the key value in the octets addressed by
.Fa sessp-\*[Gt]mackeylen .
.\"
.Pp
-Support for a specific combination of fused privacy and
+Support for a specific combination of fused privacy and
integrity-check algorithms depends on whether the underlying
hardware supports that combination.
Not all combinations are supported
by all hardware, even if the hardware supports each operation as a
stand-alone non-fused operation.
+.It Dv CIOCGSESSION2 Fa struct session2_op *sessp
+.Bd -literal
+struct session2_op {
+u_int32_t cipher; /* e.g. CRYPTO_DES_CBC */
+u_int32_t mac; /* e.g. CRYPTO_MD5_HMAC */
+
+u_int32_t keylen; /* cipher key */
+const void *key;
+int mackeylen; /* mac key */
+const void *mackey;
+
+u_int32_t ses; /* returns: ses # */
+intcrid; /* driver id + flags (rw) */
+intpad[4]; /* for future expansion */
+};
+
+.Ed
+This request is similar to CIOGSESSION except that
+.Fa
svn commit: r356870 - stable/12/usr.sbin/bhyve
Author: jhb
Date: Sat Jan 18 23:26:15 2020
New Revision: 356870
URL: https://svnweb.freebsd.org/changeset/base/356870
Log:
MFC 355634: Emulate reads of the PCI command register for passthrough devices.
VFs return zero for the memory enable bit even if it has been set by a
prior write. After r348779 this caused the annoying behavior that a
guest OS would unintentionally disable memory decoding on a future
read-modify-write operation on the command register. Instead, return
the shadow value of the command register for reads. This ensures that
the guest will only toggle the state of the memory enable bit when it
specifically intends to do so.
Sponsored by: Chelsio Communications
Modified:
stable/12/usr.sbin/bhyve/pci_passthru.c
Directory Properties:
stable/12/ (props changed)
Modified: stable/12/usr.sbin/bhyve/pci_passthru.c
==
--- stable/12/usr.sbin/bhyve/pci_passthru.c Sat Jan 18 22:58:32 2020
(r356869)
+++ stable/12/usr.sbin/bhyve/pci_passthru.c Sat Jan 18 23:26:15 2020
(r356870)
@@ -806,6 +806,19 @@ passthru_cfgread(struct vmctx *ctx, int vcpu, struct p
}
#endif
+ /*
+* Emulate the command register. If a single read reads both the
+* command and status registers, read the status register from the
+* device's config space.
+*/
+ if (coff == PCIR_COMMAND) {
+ if (bytes <= 2)
+ return (-1);
+ *rv = pci_get_cfgdata16(pi, PCIR_COMMAND) << 16 |
+ read_config(>psc_sel, PCIR_STATUS, 2);
+ return (0);
+ }
+
/* Everything else just read from the device's config space */
*rv = read_config(>psc_sel, coff, bytes);
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r356851 - stable/12
Author: jhb
Date: Fri Jan 17 22:02:35 2020
New Revision: 356851
URL: https://svnweb.freebsd.org/changeset/base/356851
Log:
MFC 353932: Strip "sf" suffix when generating a target triple.
This fixes the target triple used when compiling riscv64sf with clang.
Sponsored by: DARPA
Modified:
stable/12/Makefile.inc1
Directory Properties:
stable/12/ (props changed)
Modified: stable/12/Makefile.inc1
==
--- stable/12/Makefile.inc1 Fri Jan 17 21:57:05 2020(r356850)
+++ stable/12/Makefile.inc1 Fri Jan 17 22:02:35 2020(r356851)
@@ -126,9 +126,9 @@ TARGET_ABI= gnueabi
.endif
.endif
MACHINE_ABI?= unknown
-MACHINE_TRIPLE?=${MACHINE_ARCH:S/amd64/x86_64/:C/hf$//:S/mipsn32/mips64/}-${MACHINE_ABI}-freebsd12.1
+MACHINE_TRIPLE?=${MACHINE_ARCH:S/amd64/x86_64/:C/[hs]f$//:S/mipsn32/mips64/}-${MACHINE_ABI}-freebsd12.1
TARGET_ABI?= unknown
-TARGET_TRIPLE?=
${TARGET_ARCH:S/amd64/x86_64/:C/hf$//:S/mipsn32/mips64/}-${TARGET_ABI}-freebsd12.1
+TARGET_TRIPLE?=
${TARGET_ARCH:S/amd64/x86_64/:C/[hs]f$//:S/mipsn32/mips64/}-${TARGET_ABI}-freebsd12.1
KNOWN_ARCHES?= aarch64/arm64 \
amd64 \
arm \
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r356850 - stable/12/sys/riscv/include
Author: jhb Date: Fri Jan 17 21:57:05 2020 New Revision: 356850 URL: https://svnweb.freebsd.org/changeset/base/356850 Log: MFC 353931: Fix atomic_*cmpset32 on riscv64 with clang. The lr.w instruction used to read the value from memory sign-extends the value read from memory. GCC sign-extends the 32-bit comparison value passed in whereas clang currently does not. As a result, if the value being compared has the MSB set, the comparison fails for matching 32-bit values when compiled with clang. Use a cast to explicitly sign-extend the unsigned comparison value. This works with both GCC and clang. There is commentary in the RISC-V spec that suggests that GCC's approach is more correct, but it is not clear if the commentary in the RISC-V spec is binding. Sponsored by: DARPA Modified: stable/12/sys/riscv/include/atomic.h Directory Properties: stable/12/ (props changed) Modified: stable/12/sys/riscv/include/atomic.h == --- stable/12/sys/riscv/include/atomic.hFri Jan 17 21:39:28 2020 (r356849) +++ stable/12/sys/riscv/include/atomic.hFri Jan 17 21:57:05 2020 (r356850) @@ -116,7 +116,7 @@ atomic_cmpset_32(volatile uint32_t *p, uint32_t cmpval "bnez %1, 0b\n" "1:" : "=" (tmp), "=" (res), "+A" (*p) - : "rJ" (cmpval), "rJ" (newval) + : "rJ" ((long)(int32_t)cmpval), "rJ" (newval) : "memory"); return (!res); @@ -141,7 +141,7 @@ atomic_fcmpset_32(volatile uint32_t *p, uint32_t *cmpv "sw %0, %3\n" /* Save old value */ "2:" : "=" (tmp), "=" (res), "+A" (*p), "+A" (*cmpval) - : "rJ" (*cmpval), "rJ" (newval) + : "rJ" ((long)(int32_t)*cmpval), "rJ" (newval) : "memory"); return (!res); ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r356840 - head/sys/riscv/riscv
Author: jhb
Date: Fri Jan 17 19:13:49 2020
New Revision: 356840
URL: https://svnweb.freebsd.org/changeset/base/356840
Log:
Check for invalid sstatus values in set_mcontext().
Previously, this check was only in sys_sigreturn() which meant that
user applications could write invalid values to the register via
setcontext() or swapcontext().
Reviewed by: mhorne
Sponsored by: DARPA
Differential Revision:https://reviews.freebsd.org/D23219
Modified:
head/sys/riscv/riscv/machdep.c
Modified: head/sys/riscv/riscv/machdep.c
==
--- head/sys/riscv/riscv/machdep.c Fri Jan 17 19:01:59 2020
(r356839)
+++ head/sys/riscv/riscv/machdep.c Fri Jan 17 19:13:49 2020
(r356840)
@@ -367,6 +367,14 @@ set_mcontext(struct thread *td, mcontext_t *mcp)
tf = td->td_frame;
+ /*
+* Make sure the processor mode has not been tampered with and
+* interrupts have not been disabled.
+* Supervisor interrupts in user mode are always enabled.
+*/
+ if ((mcp->mc_gpregs.gp_sstatus & SSTATUS_SPP) != 0)
+ return (EINVAL);
+
memcpy(tf->tf_t, mcp->mc_gpregs.gp_t, sizeof(tf->tf_t));
memcpy(tf->tf_s, mcp->mc_gpregs.gp_s, sizeof(tf->tf_s));
memcpy(tf->tf_a, mcp->mc_gpregs.gp_a, sizeof(tf->tf_a));
@@ -523,21 +531,11 @@ struct sigreturn_args {
int
sys_sigreturn(struct thread *td, struct sigreturn_args *uap)
{
- uint64_t sstatus;
ucontext_t uc;
int error;
if (copyin(uap->sigcntxp, , sizeof(uc)))
return (EFAULT);
-
- /*
-* Make sure the processor mode has not been tampered with and
-* interrupts have not been disabled.
-* Supervisor interrupts in user mode are always enabled.
-*/
- sstatus = uc.uc_mcontext.mc_gpregs.gp_sstatus;
- if ((sstatus & SSTATUS_SPP) != 0)
- return (EINVAL);
error = set_mcontext(td, _mcontext);
if (error != 0)
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r356839 - in head/sys: arm64/arm64 riscv/riscv
Author: jhb
Date: Fri Jan 17 19:01:59 2020
New Revision: 356839
URL: https://svnweb.freebsd.org/changeset/base/356839
Log:
Save and restore floating point registers in get/set_mcontext().
arm64 and riscv were only saving and restoring floating point
registers for sendsig() and sys_sigreturn(), but not for getcontext(),
setcontext(), and swapcontext().
While here, remove an always-false check for uap being NULL from
sys_sigreturn().
Reviewed by: br, mhorne
Sponsored by: DARPA
Differential Revision:https://reviews.freebsd.org/D23218
Modified:
head/sys/arm64/arm64/machdep.c
head/sys/riscv/riscv/machdep.c
Modified: head/sys/arm64/arm64/machdep.c
==
--- head/sys/arm64/arm64/machdep.c Fri Jan 17 17:57:34 2020
(r356838)
+++ head/sys/arm64/arm64/machdep.c Fri Jan 17 19:01:59 2020
(r356839)
@@ -98,6 +98,8 @@ __FBSDID("$FreeBSD$");
#include
#endif
+static void get_fpcontext(struct thread *td, mcontext_t *mcp);
+static void set_fpcontext(struct thread *td, mcontext_t *mcp);
enum arm64_bus arm64_bus_method = ARM64_BUS_NONE;
@@ -473,6 +475,7 @@ get_mcontext(struct thread *td, mcontext_t *mcp, int c
mcp->mc_gpregs.gp_sp = tf->tf_sp;
mcp->mc_gpregs.gp_lr = tf->tf_lr;
mcp->mc_gpregs.gp_elr = tf->tf_elr;
+ get_fpcontext(td, mcp);
return (0);
}
@@ -495,6 +498,7 @@ set_mcontext(struct thread *td, mcontext_t *mcp)
tf->tf_lr = mcp->mc_gpregs.gp_lr;
tf->tf_elr = mcp->mc_gpregs.gp_elr;
tf->tf_spsr = mcp->mc_gpregs.gp_spsr;
+ set_fpcontext(td, mcp);
return (0);
}
@@ -667,15 +671,12 @@ sys_sigreturn(struct thread *td, struct sigreturn_args
ucontext_t uc;
int error;
- if (uap == NULL)
- return (EFAULT);
if (copyin(uap->sigcntxp, , sizeof(uc)))
return (EFAULT);
error = set_mcontext(td, _mcontext);
if (error != 0)
return (error);
- set_fpcontext(td, _mcontext);
/* Restore signal mask. */
kern_sigprocmask(td, SIG_SETMASK, _sigmask, NULL, 0);
@@ -747,7 +748,6 @@ sendsig(sig_t catcher, ksiginfo_t *ksi, sigset_t *mask
/* Fill in the frame to copy out */
bzero(, sizeof(frame));
get_mcontext(td, _uc.uc_mcontext, 0);
- get_fpcontext(td, _uc.uc_mcontext);
frame.sf_si = ksi->ksi_info;
frame.sf_uc.uc_sigmask = *mask;
frame.sf_uc.uc_stack = td->td_sigstk;
Modified: head/sys/riscv/riscv/machdep.c
==
--- head/sys/riscv/riscv/machdep.c Fri Jan 17 17:57:34 2020
(r356838)
+++ head/sys/riscv/riscv/machdep.c Fri Jan 17 19:01:59 2020
(r356839)
@@ -99,6 +99,9 @@ __FBSDID("$FreeBSD$");
#include
#endif
+static void get_fpcontext(struct thread *td, mcontext_t *mcp);
+static void set_fpcontext(struct thread *td, mcontext_t *mcp);
+
struct pcpu __pcpu[MAXCPU];
static struct trapframe proc0_tf;
@@ -352,6 +355,7 @@ get_mcontext(struct thread *td, mcontext_t *mcp, int c
mcp->mc_gpregs.gp_tp = tf->tf_tp;
mcp->mc_gpregs.gp_sepc = tf->tf_sepc;
mcp->mc_gpregs.gp_sstatus = tf->tf_sstatus;
+ get_fpcontext(td, mcp);
return (0);
}
@@ -372,6 +376,7 @@ set_mcontext(struct thread *td, mcontext_t *mcp)
tf->tf_gp = mcp->mc_gpregs.gp_gp;
tf->tf_sepc = mcp->mc_gpregs.gp_sepc;
tf->tf_sstatus = mcp->mc_gpregs.gp_sstatus;
+ set_fpcontext(td, mcp);
return (0);
}
@@ -522,8 +527,6 @@ sys_sigreturn(struct thread *td, struct sigreturn_args
ucontext_t uc;
int error;
- if (uap == NULL)
- return (EFAULT);
if (copyin(uap->sigcntxp, , sizeof(uc)))
return (EFAULT);
@@ -540,8 +543,6 @@ sys_sigreturn(struct thread *td, struct sigreturn_args
if (error != 0)
return (error);
- set_fpcontext(td, _mcontext);
-
/* Restore signal mask. */
kern_sigprocmask(td, SIG_SETMASK, _sigmask, NULL, 0);
@@ -612,7 +613,6 @@ sendsig(sig_t catcher, ksiginfo_t *ksi, sigset_t *mask
/* Fill in the frame to copy out */
bzero(, sizeof(frame));
get_mcontext(td, _uc.uc_mcontext, 0);
- get_fpcontext(td, _uc.uc_mcontext);
frame.sf_si = ksi->ksi_info;
frame.sf_uc.uc_sigmask = *mask;
frame.sf_uc.uc_stack = td->td_sigstk;
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r356831 - head/sys/powerpc/include
On 1/17/20 6:43 AM, Leandro Lupori wrote: > Author: luporl > Date: Fri Jan 17 14:43:58 2020 > New Revision: 356831 > URL: https://svnweb.freebsd.org/changeset/base/356831 > > Log: > [PPC] Fix wrong comment > > pcb_context[20] holds r12-r31 and not r14-r31, as the comment said. Thanks, this was the source of the kgdb bug I think. :) -- John Baldwin ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r356797 - head/share/mk
On 1/16/20 9:27 AM, Ed Maste wrote: > Author: emaste > Date: Thu Jan 16 17:27:08 2020 > New Revision: 356797 > URL: https://svnweb.freebsd.org/changeset/base/356797 > > Log: > pkgbase: move profiling _p.a libs into -development packages > > Profiling library archives are part of the development environment; they > don't need to be in separate -profile packages. > > (In fact we can probably just eliminate the _p.a archives assuming that > profiling will be done using hwpmc etc., but that is a change for later.) I would support having MK_PROFILE default to off for 13. WITHOUT_PROFILE=yes is something I've been adding to make.conf (or src.conf) for 2 decades. The mcount-based stuff does not seem compelling compared to sampling via hardware counters. It also requires MD code that doesn't even work on all platforms (e.g. I think the mcount bits for MIPS only work for o32 and are completely broken for n64). Do we know if any other systems still ship -pg libraries as an option? Also, is anyone still using them? -- John Baldwin ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r356755 - in head/sys: net netinet netinet6 netpfil/ipfw/nat64 sys
On 1/15/20 2:38 PM, Gleb Smirnoff wrote: > On Wed, Jan 15, 2020 at 09:44:53AM -1000, Jeff Roberson wrote: > J> On Wed, 15 Jan 2020, Gleb Smirnoff wrote: > J> > J> > Author: glebius > J> > Date: Wed Jan 15 06:05:20 2020 > J> > New Revision: 356755 > J> > URL: https://svnweb.freebsd.org/changeset/base/356755 > J> > > J> > Log: > J> > Introduce NET_EPOCH_CALL() macro and use it everywhere where we free > J> > data based on the network epoch. The macro reverses the argument > J> > order of epoch_call(9) - first function, then its argument. NFC > J> > J> Is there some practical impact of changing the argument order or does it > J> just seem more natural to you? > > It is just more natural. I'm suggesting to change prototype of epoch_call() > to the same order as well. +1 for fn, arg. -- John Baldwin ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r356737 - head/sys/mips/mips
Author: jhb
Date: Tue Jan 14 18:00:04 2020
New Revision: 356737
URL: https://svnweb.freebsd.org/changeset/base/356737
Log:
Preserve the inherited value of the status register in cpu_set_upcall().
Instead of re-deriving the value of SR using logic similar to
exec_set_regs(), just inherit the value from the existing thread
similar to fork().
Reviewed by: brooks
Obtained from:CheriBSD
Sponsored by: DARPA
Differential Revision:https://reviews.freebsd.org/D23059
Modified:
head/sys/mips/mips/vm_machdep.c
Modified: head/sys/mips/mips/vm_machdep.c
==
--- head/sys/mips/mips/vm_machdep.c Tue Jan 14 17:56:54 2020
(r356736)
+++ head/sys/mips/mips/vm_machdep.c Tue Jan 14 18:00:04 2020
(r356737)
@@ -414,7 +414,7 @@ cpu_set_upcall(struct thread *td, void (*entry)(void *
stack_t *stack)
{
struct trapframe *tf;
- register_t sp;
+ register_t sp, sr;
sp = (((intptr_t)stack->ss_sp + stack->ss_size) & ~(STACK_ALIGN - 1)) -
CALLFRAME_SIZ;
@@ -424,8 +424,10 @@ cpu_set_upcall(struct thread *td, void (*entry)(void *
* function.
*/
tf = td->td_frame;
+ sr = tf->sr;
bzero(tf, sizeof(struct trapframe));
tf->sp = sp;
+ tf->sr = sr;
tf->pc = (register_t)(intptr_t)entry;
/*
* MIPS ABI requires T9 to be the same as PC
@@ -434,18 +436,6 @@ cpu_set_upcall(struct thread *td, void (*entry)(void *
tf->t9 = (register_t)(intptr_t)entry;
tf->a0 = (register_t)(intptr_t)arg;
- /*
-* Keep interrupt mask
-*/
- td->td_frame->sr = MIPS_SR_KSU_USER | MIPS_SR_EXL | MIPS_SR_INT_IE |
- (mips_rd_status() & MIPS_SR_INT_MASK);
-#if defined(__mips_n32)
- td->td_frame->sr |= MIPS_SR_PX;
-#elif defined(__mips_n64)
- td->td_frame->sr |= MIPS_SR_PX | MIPS_SR_UX | MIPS_SR_KX;
-#endif
-/* tf->sr |= (ALL_INT_MASK & idle_mask) | SR_INT_ENAB; */
- /**XXX the above may now be wrong -- mips2 implements this as panic */
/*
* FREEBSD_DEVELOPERS_FIXME:
* Setup any other CPU-Specific registers (Not MIPS Standard)
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r356561 - head/sys/opencrypto
Author: jhb
Date: Thu Jan 9 18:29:59 2020
New Revision: 356561
URL: https://svnweb.freebsd.org/changeset/base/356561
Log:
Add stricter checking on mac key lengths.
Negative lengths are always invalid. The key length should also
be zero for hash algorithms that do not accept a key.
admbugs: 949
Reported by: Yuval Kanarenstein
Reviewed by: cem
MFC after:1 week
Sponsored by: Chelsio Communications
Differential Revision:https://reviews.freebsd.org/D23094
Modified:
head/sys/opencrypto/cryptodev.c
Modified: head/sys/opencrypto/cryptodev.c
==
--- head/sys/opencrypto/cryptodev.c Thu Jan 9 18:14:48 2020
(r356560)
+++ head/sys/opencrypto/cryptodev.c Thu Jan 9 18:29:59 2020
(r356561)
@@ -585,8 +585,8 @@ cryptof_ioctl(
if (thash) {
cria.cri_alg = thash->type;
cria.cri_klen = sop->mackeylen * 8;
- if (thash->keysize != 0 &&
- sop->mackeylen > thash->keysize) {
+ if (sop->mackeylen > thash->keysize ||
+ sop->mackeylen < 0) {
CRYPTDEB("invalid mac key length");
error = EINVAL;
SDT_PROBE1(opencrypto, dev, ioctl, error,
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r356520 - head/sys/opencrypto
Author: jhb
Date: Wed Jan 8 22:16:26 2020
New Revision: 356520
URL: https://svnweb.freebsd.org/changeset/base/356520
Log:
Remove no-longer-used function prototype.
Reported by: amd64-gcc
Modified:
head/sys/opencrypto/cryptodev.c
Modified: head/sys/opencrypto/cryptodev.c
==
--- head/sys/opencrypto/cryptodev.c Wed Jan 8 22:06:31 2020
(r356519)
+++ head/sys/opencrypto/cryptodev.c Wed Jan 8 22:16:26 2020
(r356520)
@@ -326,7 +326,6 @@ static struct fileops cryptofops = {
static struct csession *csefind(struct fcrypt *, u_int);
static bool csedelete(struct fcrypt *, u_int);
-static void cseadd(struct fcrypt *, struct csession *);
static struct csession *csecreate(struct fcrypt *, crypto_session_t, caddr_t,
u_int64_t, caddr_t, u_int64_t, u_int32_t, u_int32_t, struct enc_xform *,
struct auth_hash *);
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r356509 - head/sys/opencrypto
Author: jhb
Date: Wed Jan 8 19:05:23 2020
New Revision: 356509
URL: https://svnweb.freebsd.org/changeset/base/356509
Log:
Remove unneeded cdevsw methods and D_NEEDGIANT.
Reviewed by: kib
MFC after:2 weeks
Sponsored by: Chelsio Communications
Differential Revision:https://reviews.freebsd.org/D23079
Modified:
head/sys/opencrypto/cryptodev.c
Modified: head/sys/opencrypto/cryptodev.c
==
--- head/sys/opencrypto/cryptodev.c Wed Jan 8 19:03:24 2020
(r356508)
+++ head/sys/opencrypto/cryptodev.c Wed Jan 8 19:05:23 2020
(r356509)
@@ -1473,24 +1473,6 @@ csefree(struct csession *cse)
}
static int
-cryptoopen(struct cdev *dev, int oflags, int devtype, struct thread *td)
-{
- return (0);
-}
-
-static int
-cryptoread(struct cdev *dev, struct uio *uio, int ioflag)
-{
- return (EIO);
-}
-
-static int
-cryptowrite(struct cdev *dev, struct uio *uio, int ioflag)
-{
- return (EIO);
-}
-
-static int
cryptoioctl(struct cdev *dev, u_long cmd, caddr_t data, int flag, struct
thread *td)
{
struct file *f;
@@ -1531,10 +1513,6 @@ cryptoioctl(struct cdev *dev, u_long cmd, caddr_t data
static struct cdevsw crypto_cdevsw = {
.d_version =D_VERSION,
- .d_flags = D_NEEDGIANT,
- .d_open = cryptoopen,
- .d_read = cryptoread,
- .d_write = cryptowrite,
.d_ioctl = cryptoioctl,
.d_name = "crypto",
};
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r356508 - head/sys/opencrypto
Author: jhb
Date: Wed Jan 8 19:03:24 2020
New Revision: 356508
URL: https://svnweb.freebsd.org/changeset/base/356508
Log:
Use falloc_noinstall + finstall for crypto file descriptors.
Reviewed by: cem, kib
MFC after:1 week
Sponsored by: Chelsio Communications
Differential Revision:https://reviews.freebsd.org/D23078
Modified:
head/sys/opencrypto/cryptodev.c
Modified: head/sys/opencrypto/cryptodev.c
==
--- head/sys/opencrypto/cryptodev.c Wed Jan 8 18:59:23 2020
(r356507)
+++ head/sys/opencrypto/cryptodev.c Wed Jan 8 19:03:24 2020
(r356508)
@@ -1499,20 +1499,21 @@ cryptoioctl(struct cdev *dev, u_long cmd, caddr_t data
switch (cmd) {
case CRIOGET:
+ error = falloc_noinstall(td, );
+ if (error)
+ break;
+
fcr = malloc(sizeof(struct fcrypt), M_XDATA, M_WAITOK | M_ZERO);
TAILQ_INIT(>csessions);
mtx_init(>lock, "fcrypt", NULL, MTX_DEF);
- error = falloc(td, , , 0);
-
+ finit(f, FREAD | FWRITE, DTYPE_CRYPTO, fcr, );
+ error = finstall(td, f, , 0, NULL);
if (error) {
mtx_destroy(>lock);
free(fcr, M_XDATA);
- return (error);
- }
- /* falloc automatically provides an extra reference to 'f'. */
- finit(f, FREAD | FWRITE, DTYPE_CRYPTO, fcr, );
- *(u_int32_t *)data = fd;
+ } else
+ *(uint32_t *)data = fd;
fdrop(f, td);
break;
case CRIOFINDDEV:
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r356507 - head/sys/opencrypto
Author: jhb
Date: Wed Jan 8 18:59:23 2020
New Revision: 356507
URL: https://svnweb.freebsd.org/changeset/base/356507
Log:
Add a reference count to cryptodev sessions.
This prevents use-after-free races with crypto requests (which may
sleep) and CIOCFSESSION as well as races from current CIOCFSESSION
requests.
admbugs: 949
Reported by: Yuval Kanarenstein
Reviewed by: cem
MFC after:1 week
Sponsored by: Chelsio Communications
Differential Revision:https://reviews.freebsd.org/D23077
Modified:
head/sys/opencrypto/cryptodev.c
Modified: head/sys/opencrypto/cryptodev.c
==
--- head/sys/opencrypto/cryptodev.c Wed Jan 8 18:26:23 2020
(r356506)
+++ head/sys/opencrypto/cryptodev.c Wed Jan 8 18:59:23 2020
(r356507)
@@ -266,6 +266,7 @@ crypt_kop_to_32(const struct crypt_kop *from, struct c
struct csession {
TAILQ_ENTRY(csession) next;
crypto_session_t cses;
+ volatile u_int refs;
u_int32_t ses;
struct mtx lock; /* for op submission */
@@ -292,6 +293,7 @@ struct cryptop_data {
struct fcrypt {
TAILQ_HEAD(csessionlist, csession) csessions;
int sesn;
+ struct mtx lock;
};
static struct timeval warninterval = { .tv_sec = 60, .tv_usec = 0 };
@@ -323,8 +325,8 @@ static struct fileops cryptofops = {
};
static struct csession *csefind(struct fcrypt *, u_int);
-static int csedelete(struct fcrypt *, struct csession *);
-static struct csession *cseadd(struct fcrypt *, struct csession *);
+static bool csedelete(struct fcrypt *, u_int);
+static void cseadd(struct fcrypt *, struct csession *);
static struct csession *csecreate(struct fcrypt *, crypto_session_t, caddr_t,
u_int64_t, caddr_t, u_int64_t, u_int32_t, u_int32_t, struct enc_xform *,
struct auth_hash *);
@@ -668,13 +670,10 @@ bail:
break;
case CIOCFSESSION:
ses = *(u_int32_t *)data;
- cse = csefind(fcr, ses);
- if (cse == NULL) {
+ if (!csedelete(fcr, ses)) {
SDT_PROBE1(opencrypto, dev, ioctl, error, __LINE__);
return (EINVAL);
}
- csedelete(fcr, cse);
- csefree(cse);
break;
case CIOCCRYPT:
#ifdef COMPAT_FREEBSD32
@@ -691,6 +690,7 @@ bail:
return (EINVAL);
}
error = cryptodev_op(cse, cop, active_cred, td);
+ csefree(cse);
#ifdef COMPAT_FREEBSD32
if (error == 0 && cmd == CIOCCRYPT32)
crypt_op_to_32(cop, data);
@@ -757,6 +757,7 @@ bail:
return (EINVAL);
}
error = cryptodev_aead(cse, caead, active_cred, td);
+ csefree(cse);
break;
default:
error = EINVAL;
@@ -1375,6 +1376,9 @@ cryptof_close(struct file *fp, struct thread *td)
while ((cse = TAILQ_FIRST(>csessions))) {
TAILQ_REMOVE(>csessions, cse, next);
+ KASSERT(cse->refs == 1,
+ ("%s: crypto session %p with %d refs", __func__, cse,
+ cse->refs));
csefree(cse);
}
free(fcr, M_XDATA);
@@ -1395,34 +1399,36 @@ csefind(struct fcrypt *fcr, u_int ses)
{
struct csession *cse;
- TAILQ_FOREACH(cse, >csessions, next)
- if (cse->ses == ses)
+ mtx_lock(>lock);
+ TAILQ_FOREACH(cse, >csessions, next) {
+ if (cse->ses == ses) {
+ refcount_acquire(>refs);
+ mtx_unlock(>lock);
return (cse);
+ }
+ }
+ mtx_unlock(>lock);
return (NULL);
}
-static int
-csedelete(struct fcrypt *fcr, struct csession *cse_del)
+static bool
+csedelete(struct fcrypt *fcr, u_int ses)
{
struct csession *cse;
+ mtx_lock(>lock);
TAILQ_FOREACH(cse, >csessions, next) {
- if (cse == cse_del) {
+ if (cse->ses == ses) {
TAILQ_REMOVE(>csessions, cse, next);
- return (1);
+ mtx_unlock(>lock);
+ csefree(cse);
+ return (true);
}
}
- return (0);
+ mtx_unlock(>lock);
+ return (false);
}
-static struct csession *
-cseadd(struct fcrypt *fcr, struct csession *cse)
-{
- TAILQ_INSERT_TAIL(>csessions, cse, next);
- cse->ses = fcr->sesn++;
- return (cse);
-}
-
struct csession *
csecreate(struct fcrypt *fcr, crypto_session_t cses, caddr_t key, u_int64_t
keylen,
caddr_t mackey, u_int64_t mackeylen, u_int32_t cipher, u_int32_t mac,
@@ -1434,6 +1440,7 @@ csecreate(struct fcrypt *fcr, crypto_session_t cses, c
svn commit: r356504 - head/share/mk
Author: jhb
Date: Wed Jan 8 17:49:34 2020
New Revision: 356504
URL: https://svnweb.freebsd.org/changeset/base/356504
Log:
Add -mno-relax to CFLAGS in bsd.prog/lib.mk instead of bsd.cpu.mk.
bsd.cpu.mk is included by bsd.init.mk before bsd.linker.mk, so it
was always setting the flag since LINKER_FEATURES wasn't defined.
Reported by: mhorne
Reviewed by: imp, mhorne
Sponsored by: DARPA
Differential Revision:https://reviews.freebsd.org/D23076
Modified:
head/share/mk/bsd.cpu.mk
head/share/mk/bsd.lib.mk
head/share/mk/bsd.prog.mk
Modified: head/share/mk/bsd.cpu.mk
==
--- head/share/mk/bsd.cpu.mkWed Jan 8 17:37:20 2020(r356503)
+++ head/share/mk/bsd.cpu.mkWed Jan 8 17:49:34 2020(r356504)
@@ -375,10 +375,6 @@ CFLAGS += -march=rv64imac -mabi=lp64
.else
CFLAGS += -march=rv64imafdc -mabi=lp64d
.endif
-
-.if ${LINKER_FEATURES:U:Mriscv-relaxations} == ""
-CFLAGS += -mno-relax
-.endif
.endif
# NB: COPTFLAGS is handled in /usr/src/sys/conf/kern.pre.mk
Modified: head/share/mk/bsd.lib.mk
==
--- head/share/mk/bsd.lib.mkWed Jan 8 17:37:20 2020(r356503)
+++ head/share/mk/bsd.lib.mkWed Jan 8 17:49:34 2020(r356504)
@@ -98,6 +98,10 @@ STATIC_CFLAGS+= -ftls-model=initial-exec
STATIC_CXXFLAGS+= -ftls-model=initial-exec
.endif
+.if ${MACHINE_CPUARCH} == "riscv" && ${LINKER_FEATURES:Mriscv-relaxations} ==
""
+CFLAGS += -mno-relax
+.endif
+
.include
# prefer .s to a .c, add .po, remove stuff not used in the BSD libraries
Modified: head/share/mk/bsd.prog.mk
==
--- head/share/mk/bsd.prog.mk Wed Jan 8 17:37:20 2020(r356503)
+++ head/share/mk/bsd.prog.mk Wed Jan 8 17:49:34 2020(r356504)
@@ -57,6 +57,10 @@ LDFLAGS+= -Wl,-zretpolineplt
.endif
.endif
+.if ${MACHINE_CPUARCH} == "riscv" && ${LINKER_FEATURES:Mriscv-relaxations} ==
""
+CFLAGS += -mno-relax
+.endif
+
.if defined(CRUNCH_CFLAGS)
CFLAGS+=${CRUNCH_CFLAGS}
.else
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r356502 - head
Author: jhb Date: Wed Jan 8 17:31:54 2020 New Revision: 356502 URL: https://svnweb.freebsd.org/changeset/base/356502 Log: Add notes for MAKE_OBSOLETE_GCC going away and riscv switching to clang/lld. Modified: head/UPDATING Modified: head/UPDATING == --- head/UPDATING Wed Jan 8 17:31:18 2020(r356501) +++ head/UPDATING Wed Jan 8 17:31:54 2020(r356502) @@ -26,6 +26,15 @@ NOTE TO PEOPLE WHO THINK THAT FreeBSD 13.x IS SLOW: disable the most expensive debugging functionality run "ln -s 'abort:false,junk:false' /etc/malloc.conf".) +20200108: + Clang/LLVM is now the default compiler and LLD the default + linker for riscv64. + +20200107: + make universe no longer uses GCC 4.2.1 on any architectures. + Architectures not supported by in-tree Clang/LLVM require an + external toolchain package. + 20200104: GCC 4.2.1 is now not built by default, as part of the GCC 4.2.1 retirement plan. Specifically, the GCC, GCC_BOOTSTRAP, and GNUCXX ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r356501 - head/share/man/man7
Author: jhb Date: Wed Jan 8 17:31:18 2020 New Revision: 356501 URL: https://svnweb.freebsd.org/changeset/base/356501 Log: Remove a mention of MAKE_OBSOLETE_GCC. This should have been included in r356452. Modified: head/share/man/man7/arch.7 Modified: head/share/man/man7/arch.7 == --- head/share/man/man7/arch.7 Wed Jan 8 17:30:14 2020(r356500) +++ head/share/man/man7/arch.7 Wed Jan 8 17:31:18 2020(r356501) @@ -330,9 +330,7 @@ Note that GCC 4.2.1 is deprecated, and scheduled for r Any CPU architectures not migrated by then (to either base system Clang or external toolchain) may be removed from the tree after that date. -Unless the make variable -.Dv MAKE_OBSOLETE_GCC -is defined, make universe will not build mips or sparc64 +make universe will not build mips or sparc64 architectures unless the xtoolchain binaries have been installed for the architecture. .Ss MACHINE_ARCH vs MACHINE_CPUARCH vs MACHINE ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r356499 - in head: . share/man/man7 share/mk
Author: jhb
Date: Wed Jan 8 17:25:59 2020
New Revision: 356499
URL: https://svnweb.freebsd.org/changeset/base/356499
Log:
Use clang and lld as the default toolchain for RISCV.
- Enable clang and lld as system toolchains.
- Don't use external GCC for universe by default.
- Re-enable riscv64sf since it builds fine with clang + lld.
Reviewed by: emaste, mhorne
Relnotes: yes
Sponsored by: DARPA
Differential Revision:https://reviews.freebsd.org/D23089
Modified:
head/Makefile
head/share/man/man7/arch.7
head/share/mk/src.opts.mk
Modified: head/Makefile
==
--- head/Makefile Wed Jan 8 17:23:52 2020(r356498)
+++ head/Makefile Wed Jan 8 17:25:59 2020(r356499)
@@ -495,23 +495,20 @@ TARGET_ARCHES_arm64?= aarch64
TARGET_ARCHES_mips?= mipsel mips mips64el mips64 mipsn32 mipselhf mipshf
mips64elhf mips64hf
# powerpcspe excluded until clang fixed
TARGET_ARCHES_powerpc?=powerpc powerpc64
-# riscv64sf excluded due to PR 232085
-TARGET_ARCHES_riscv?= riscv64
+TARGET_ARCHES_riscv?= riscv64 riscv64sf
.for target in ${TARGETS}
TARGET_ARCHES_${target}?= ${target}
.endfor
-MAKE_PARAMS_riscv?=CROSS_TOOLCHAIN=riscv64-gcc
MAKE_PARAMS_mips?= CROSS_TOOLCHAIN=mips-gcc6
MAKE_PARAMS_sparc64?= CROSS_TOOLCHAIN=sparc64-gcc6
TOOLCHAINS_mips= mips-gcc6
-TOOLCHAINS_riscv= riscv64-gcc
TOOLCHAINS_sparc64=sparc64-gcc6
# Remove architectures only supported by external toolchain from
# universe if required toolchain packages are missing.
-.for target in mips riscv sparc64
+.for target in mips sparc64
.if ${_UNIVERSE_TARGETS:M${target}}
.for toolchain in ${TOOLCHAINS_${target}}
.if !exists(/usr/local/share/toolchains/${toolchain}.mk)
Modified: head/share/man/man7/arch.7
==
--- head/share/man/man7/arch.7 Wed Jan 8 17:23:52 2020(r356498)
+++ head/share/man/man7/arch.7 Wed Jan 8 17:25:59 2020(r356499)
@@ -26,7 +26,7 @@
.\"
.\" $FreeBSD$
.\"
-.Dd January 5, 2020
+.Dd January 8, 2020
.Dt ARCH 7
.Os
.Sh NAME
@@ -319,8 +319,8 @@ This table shows the default tool chain for each archi
.It powerpc Ta Clang Ta GNU ld 2.17.50
.It powerpcspe Ta Clang Ta GNU ld 2.17.50
.It powerpc64 Ta Clang Ta lld
-.It riscv64 Ta GCC(1)Ta GNU ld(1)
-.It riscv64sf Ta GCC(1)Ta GNU ld(1)
+.It riscv64 Ta Clang Ta lld
+.It riscv64sf Ta Clang Ta lld
.It sparc64 Ta GCC(1)Ta GNU ld(1)
.El
.Pp
Modified: head/share/mk/src.opts.mk
==
--- head/share/mk/src.opts.mk Wed Jan 8 17:23:52 2020(r356498)
+++ head/share/mk/src.opts.mk Wed Jan 8 17:25:59 2020(r356499)
@@ -295,8 +295,7 @@ __DEFAULT_NO_OPTIONS+=LLVM_TARGET_BPF
# If the compiler is not C++11 capable, disable Clang. External toolchain will
# be required.
-.if ${COMPILER_FEATURES:Mc++11} && (${__TT} != "mips" && \
-${__TT} != "riscv" && ${__TT} != "sparc64")
+.if ${COMPILER_FEATURES:Mc++11} && (${__TT} != "mips" && ${__TT} != "sparc64")
# Clang is enabled, and will be installed as the default /usr/bin/cc.
__DEFAULT_YES_OPTIONS+=CLANG CLANG_BOOTSTRAP CLANG_IS_CC LLD
.elif ${COMPILER_FEATURES:Mc++11} && ${__T} != "sparc64"
@@ -323,7 +322,7 @@ __DEFAULT_YES_OPTIONS+=LLVM_LIBUNWIND
__DEFAULT_NO_OPTIONS+=LLVM_LIBUNWIND
.endif
.if ${__TT} != "mips" && ${__T} != "powerpc" && ${__T} != "powerpcspe" && \
-${__TT} != "riscv" && ${__T} != "sparc64"
+${__T} != "sparc64"
__DEFAULT_YES_OPTIONS+=LLD_BOOTSTRAP LLD_IS_LD
.else
__DEFAULT_NO_OPTIONS+=LLD_BOOTSTRAP LLD_IS_LD
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r356481 - in head/sys: conf riscv/riscv
Author: jhb
Date: Tue Jan 7 23:18:31 2020
New Revision: 356481
URL: https://svnweb.freebsd.org/changeset/base/356481
Log:
Work around lld's inability to handle undefined weak symbols on risc-v.
lld on RISC-V is not yet able to handle undefined weak symbols for
non-PIC code in the code model (medany/medium) used by the RISC-V
kernel.
Both GCC and clang emit an auipc / addi pair of instructions to
generate an address relative to the current PC with a 31-bit offset.
Undefined weak symbols need to have an address of 0, but the kernel
runs with PC values much greater than 2^31, so there is no way to
construct a NULL pointer as a PC-relative value. The bfd linker
rewrites the instruction pair to use lui / addi with values of 0 to
force a NULL pointer address. (There are similar cases for 'ld'
becoming auipc / ld that bfd rewrites to lui / ld with an address of
0.)
To work around this, compile the kernel with -fPIE when using lld.
This does not make the kernel position-independent, but it does
force the compiler to indirect address lookups through GOT entries
(so auipc / ld against a GOT entry to fetch the address). This
adds extra memory indirections for global symbols, so should be
disabled once lld is finally fixed.
A few 'la' instructions in locore that depend on PC-relative
addressing to load physical addresses before paging is enabled have to
use auipc / addi and not indirect via GOT entries, so change those to
use 'lla' which always uses auipc / addi for both PIC and non-PIC.
Submitted by: jrtc27
Sponsored by: DARPA
Differential Revision:https://reviews.freebsd.org/D23064
Modified:
head/sys/conf/kern.pre.mk
head/sys/riscv/riscv/locore.S
Modified: head/sys/conf/kern.pre.mk
==
--- head/sys/conf/kern.pre.mk Tue Jan 7 23:10:38 2020(r356480)
+++ head/sys/conf/kern.pre.mk Tue Jan 7 23:18:31 2020(r356481)
@@ -170,6 +170,17 @@ LDFLAGS+= -z notext -z ifunc-noplt
.endif
.endif
+.if ${MACHINE_CPUARCH} == "riscv"
+# Hack: Work around undefined weak symbols being out of range when linking with
+# LLD (address is a PC-relative calculation, and BFD works around this by
+# rewriting the instructions to generate an absolute address of 0); -fPIE
+# avoids this since it uses the GOT for all extern symbols, which is overly
+# inefficient for us. Drop once undefined weak symbols work with medany.
+.if ${LINKER_TYPE} == "lld"
+CFLAGS+= -fPIE
+.endif
+.endif
+
NORMAL_C= ${CC} -c ${CFLAGS} ${WERROR} ${PROF} ${.IMPSRC}
NORMAL_S= ${CC:N${CCACHE_BIN}} -c ${ASM_CFLAGS} ${WERROR} ${.IMPSRC}
PROFILE_C= ${CC} -c ${CFLAGS} ${WERROR} ${.IMPSRC}
Modified: head/sys/riscv/riscv/locore.S
==
--- head/sys/riscv/riscv/locore.S Tue Jan 7 23:10:38 2020
(r356480)
+++ head/sys/riscv/riscv/locore.S Tue Jan 7 23:18:31 2020
(r356481)
@@ -54,7 +54,7 @@
.globl _start
_start:
/* Get the physical address kernel loaded to */
- la t0, virt_map
+ lla t0, virt_map
ld t1, 0(t0)
sub t1, t1, t0
li t2, KERNBASE
@@ -66,7 +66,7 @@ _start:
*/
/* Pick a hart to run the boot process. */
- la t0, hart_lottery
+ lla t0, hart_lottery
li t1, 1
amoadd.w t0, t1, 0(t0)
@@ -82,8 +82,8 @@ _start:
*/
1:
/* Add L1 entry for kernel */
- la s1, pagetable_l1
- la s2, pagetable_l2/* Link to next level PN */
+ lla s1, pagetable_l1
+ lla s2, pagetable_l2/* Link to next level PN */
srlis2, s2, PAGE_SHIFT
li a5, KERNBASE
@@ -100,7 +100,7 @@ _start:
sd t6, (t0)
/* Level 2 superpages (512 x 2MiB) */
- la s1, pagetable_l2
+ lla s1, pagetable_l2
srlit4, s9, 21 /* Div physmem base by 2 MiB */
li t2, 512 /* Build 512 entries */
add t3, t4, t2
@@ -116,8 +116,8 @@ _start:
bltut4, t3, 2b
/* Create an L1 page for early devmap */
- la s1, pagetable_l1
- la s2, pagetable_l2_devmap /* Link to next level PN */
+ lla s1, pagetable_l1
+ lla s2, pagetable_l2_devmap /* Link to next level PN */
srlis2, s2, PAGE_SHIFT
li a5, (VM_MAX_KERNEL_ADDRESS - L2_SIZE)
@@ -134,7 +134,7 @@ _start:
sd t6, (t0)
/* Create an L2 page superpage for DTB */
- la s1, pagetable_l2_devmap
+ lla s1, pagetable_l2_devmap
mv s2, a1
srlis2, s2, PAGE_SHIFT
@@ -152,14 +152,14 @@ _start:
/* Page tables END */
/* Setup supervisor trap vector */
- la t0, va
+ lla t0, va
sub t0, t0, s9
svn commit: r356478 - head
Author: jhb
Date: Tue Jan 7 21:56:28 2020
New Revision: 356478
URL: https://svnweb.freebsd.org/changeset/base/356478
Log:
Don't fail universe kernel stage for TARGET_ARCHes without a kernel config.
This fixes a regression in r356418 where the entire universe would
fail early due to an undefined make target when a given TARGET_ARCH
had no associated kernel configs. This is true for all of the
hard-float mips TARGET_ARCHes currently.
Pointy hat to:me
Reviewed by: emaste
MFC after:1 month
Differential Revision:https://reviews.freebsd.org/D23071
Modified:
head/Makefile
Modified: head/Makefile
==
--- head/Makefile Tue Jan 7 21:56:20 2020(r356477)
+++ head/Makefile Tue Jan 7 21:56:28 2020(r356478)
@@ -728,6 +728,7 @@ universe_kernconf_${TARGET}_${kernel}: .MAKE
.endfor
.for target_arch in ${TARGET_ARCHES_${TARGET}}
universe_kernconfs: universe_kernconfs_${target_arch} .PHONY
+universe_kernconfs_${target_arch}:
.endfor
.endif # make(universe_kernels)
universe: universe_epilogue
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r356457 - head/tools/tools/crypto
Author: jhb
Date: Tue Jan 7 17:07:58 2020
New Revision: 356457
URL: https://svnweb.freebsd.org/changeset/base/356457
Log:
Various cleanups to cryptocheck.
- Rename 'blkcipher' to 'cipher'. Some of the ciphers being tested
are stream ciphers.
- Rename 'authenc' to 'eta' as it is only testing ETA chained
operations and not other combination modes.
- Add a notion of an OCF session and some helper routines to try to
reduce duplicated code. This also uses a single session for both
encrypt and decrypt operations during a single test.
- Add tests to ensure that AEAD algorithms fail decryption with
EBADMSG when given a corrupted tag.
- Remove the transitional hack for COP_F_CIPHER_FIRST.
- Update block comment to mention plain hashes.
Reviewed by: cem
Sponsored by: Chelsio Communications
Differential Revision:https://reviews.freebsd.org/D22940
Modified:
head/tools/tools/crypto/cryptocheck.c
Modified: head/tools/tools/crypto/cryptocheck.c
==
--- head/tools/tools/crypto/cryptocheck.c Tue Jan 7 17:02:49 2020
(r356456)
+++ head/tools/tools/crypto/cryptocheck.c Tue Jan 7 17:07:58 2020
(r356457)
@@ -73,41 +73,50 @@
*
* Supported algorithms:
* all Run all tests
- * hmacRun all hmac tests
- * blkcipher Run all block cipher tests
- * authenc Run all authenticated encryption tests
+ * hashRun all hash tests
+ * mac Run all mac tests
+ * cipher Run all cipher tests
+ * eta Run all encrypt-then-authenticate tests
* aeadRun all authenticated encryption with associated data
* tests
*
- * HMACs:
- * sha1sha1 hmac
- * sha256 256-bit sha2 hmac
- * sha384 384-bit sha2 hmac
- * sha512 512-bit sha2 hmac
+ * Hashes:
+ * sha1SHA-1
+ * sha224 224-bit SHA-2
+ * sha256 256-bit SHA-2
+ * sha384 384-bit SHA-2
+ * sha512 512-bit SHA-2
* blake2b Blake2-B
* blake2s Blake2-S
*
- * Block Ciphers:
- * aes-cbc 128-bit aes cbc
- * aes-cbc192 192-bit aes cbc
- * aes-cbc256 256-bit aes cbc
- * aes-ctr 128-bit aes ctr
- * aes-ctr192 192-bit aes ctr
- * aes-ctr256 256-bit aes ctr
- * aes-xts 128-bit aes xts
- * aes-xts256 256-bit aes xts
+ * MACs:
+ * sha1hmacSHA-1 HMAC
+ * sha224hmac 224-bit SHA-2 HMAC
+ * sha256hmac 256-bit SHA-2 HMAC
+ * sha384hmac 384-bit SHA-2 HMAC
+ * sha512hmac 512-bit SHA-2 HMAC
+ *
+ * Ciphers:
+ * aes-cbc 128-bit AES-CBC
+ * aes-cbc192 192-bit AES-CBC
+ * aes-cbc256 256-bit AES-CBC
+ * aes-ctr 128-bit AES-CTR
+ * aes-ctr192 192-bit AES-CTR
+ * aes-ctr256 256-bit AES-CTR
+ * aes-xts 128-bit AES-XTS
+ * aes-xts256 256-bit AES-XTS
* chacha20
*
- * Authenticated Encryption:
- * +
+ * Encrypt then Authenticate:
+ * +
*
* Authenticated Encryption with Associated Data:
- * aes-gcm 128-bit aes gcm
- * aes-gcm192 192-bit aes gcm
- * aes-gcm256 256-bit aes gcm
- * aes-ccm 128-bit aes ccm
- * aes-ccm192 192-bit aes ccm
- * aes-ccm256 256-bit aes ccm
+ * aes-gcm 128-bit AES-GCM
+ * aes-gcm192 192-bit AES-GCM
+ * aes-gcm256 256-bit AES-GCM
+ * aes-ccm 128-bit AES-CCM
+ * aes-ccm192 192-bit AES-CCM
+ * aes-ccm256 256-bit AES-CCM
*/
#include
@@ -126,16 +135,17 @@
#include
-/* XXX: Temporary hack */
-#ifndef COP_F_CIPHER_FIRST
-#defineCOP_F_CIPHER_FIRST 0x0001 /* Cipher before MAC. */
-#endif
+struct ocf_session {
+ int fd;
+ int ses;
+ int crid;
+};
-struct alg {
+const struct alg {
const char *name;
int cipher;
int mac;
- enum { T_HASH, T_HMAC, T_BLKCIPHER, T_AUTHENC, T_GCM, T_CCM } type;
+ enum { T_HASH, T_HMAC, T_CIPHER, T_ETA, T_AEAD } type;
const EVP_CIPHER *(*evp_cipher)(void);
const EVP_MD *(*evp_md)(void);
} algs[] = {
@@ -163,41 +173,41 @@ struct alg {
.evp_md = EVP_blake2b512 },
{ .name = "blake2s", .mac = CRYPTO_BLAKE2S, .type = T_HASH,
.evp_md = EVP_blake2s256 },
- { .name = "aes-cbc", .cipher = CRYPTO_AES_CBC, .type = T_BLKCIPHER,
+ { .name = "aes-cbc", .cipher = CRYPTO_AES_CBC, .type = T_CIPHER,
.evp_cipher = EVP_aes_128_cbc },
- { .name = "aes-cbc192", .cipher = CRYPTO_AES_CBC, .type = T_BLKCIPHER,
+ { .name = "aes-cbc192", .cipher = CRYPTO_AES_CBC, .type = T_CIPHER,
.evp_cipher = EVP_aes_192_cbc },
- { .name = "aes-cbc256", .cipher
svn commit: r356452 - head
Author: jhb
Date: Tue Jan 7 16:53:51 2020
New Revision: 356452
URL: https://svnweb.freebsd.org/changeset/base/356452
Log:
Remove support for using GCC 4.2.1 from make tinderbox.
Reviewed by: emaste
Differential Revision:https://reviews.freebsd.org/D23054
Modified:
head/Makefile
Modified: head/Makefile
==
--- head/Makefile Tue Jan 7 16:52:15 2020(r356451)
+++ head/Makefile Tue Jan 7 16:53:51 2020(r356452)
@@ -6,9 +6,7 @@
# universe- *Really* build *everything* (buildworld and
# all kernels on all architectures). Define
# MAKE_JUST_KERNELS to only build kernels,
-# MAKE_JUST_WORLDS to only build userland, and/or
-# MAKE_OBSOLETE_GCC to also build architectures
-# unsupported by clang using in-tree gcc.
+# MAKE_JUST_WORLDS to only build userland.
# tinderbox - Same as universe, but presents a list of failed build
# targets and exits with an error if there were any.
# buildworld - Rebuild *everything*, including glue to help do
@@ -490,8 +488,7 @@ worlds: .PHONY
# In all cases, if the user specifies TARGETS on the command line,
# honor that most of all.
#
-_OBSOLETE_GCC_TARGETS=mips sparc64
-TARGETS?=amd64 arm arm64 i386 powerpc riscv ${_OBSOLETE_GCC_TARGETS}
+TARGETS?=amd64 arm arm64 i386 mips powerpc riscv sparc64
_UNIVERSE_TARGETS= ${TARGETS}
TARGET_ARCHES_arm?=armv6 armv7
TARGET_ARCHES_arm64?= aarch64
@@ -505,21 +502,16 @@ TARGET_ARCHES_${target}?= ${target}
.endfor
MAKE_PARAMS_riscv?=CROSS_TOOLCHAIN=riscv64-gcc
-.if !defined(MAKE_OBSOLETE_GCC)
-OBSOLETE_GCC_TARGETS=${_OBSOLETE_GCC_TARGETS}
MAKE_PARAMS_mips?= CROSS_TOOLCHAIN=mips-gcc6
MAKE_PARAMS_sparc64?= CROSS_TOOLCHAIN=sparc64-gcc6
-.endif
TOOLCHAINS_mips= mips-gcc6
TOOLCHAINS_riscv= riscv64-gcc
TOOLCHAINS_sparc64=sparc64-gcc6
# Remove architectures only supported by external toolchain from
-# universe if required toolchain packages are missing. riscv requires
-# an out-of-tree toolchain. When MAKE_OBSOLETE_GCC is not defined,
-# the same logic appleis to the obsolete gcc targets.
-.for target in riscv ${OBSOLETE_GCC_TARGETS}
+# universe if required toolchain packages are missing.
+.for target in mips riscv sparc64
.if ${_UNIVERSE_TARGETS:M${target}}
.for toolchain in ${TOOLCHAINS_${target}}
.if !exists(/usr/local/share/toolchains/${toolchain}.mk)
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r356420 - head/sys/mips/mips
Author: jhb
Date: Mon Jan 6 18:02:02 2020
New Revision: 356420
URL: https://svnweb.freebsd.org/changeset/base/356420
Log:
Simplify arguments to signal handlers on mips.
- Use ksi_addr directly as si_addr in the siginfo instead of the
'badvaddr' register.
- Remove a duplicate assignment of si_code.
- Use ksi_addr as the 4th argument to the old-style handler instead of
'badvaddr'.
Reviewed by: brooks, kevans
Sponsored by: DARPA
Differential Revision:https://reviews.freebsd.org/D23013
Modified:
head/sys/mips/mips/pm_machdep.c
Modified: head/sys/mips/mips/pm_machdep.c
==
--- head/sys/mips/mips/pm_machdep.c Mon Jan 6 17:36:28 2020
(r356419)
+++ head/sys/mips/mips/pm_machdep.c Mon Jan 6 18:02:02 2020
(r356420)
@@ -147,12 +147,10 @@ sendsig(sig_t catcher, ksiginfo_t *ksi, sigset_t *mask
/* fill siginfo structure */
sf.sf_si = ksi->ksi_info;
sf.sf_si.si_signo = sig;
- sf.sf_si.si_code = ksi->ksi_code;
- sf.sf_si.si_addr = (void*)(intptr_t)regs->badvaddr;
} else {
/* Old FreeBSD-style arguments. */
regs->a1 = ksi->ksi_code;
- regs->a3 = regs->badvaddr;
+ regs->a3 = (uintptr_t)ksi->ksi_addr;
/* sf.sf_ahu.sf_handler = catcher; */
}
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r356419 - head
Author: jhb
Date: Mon Jan 6 17:36:28 2020
New Revision: 356419
URL: https://svnweb.freebsd.org/changeset/base/356419
Log:
Enable powerpc in make tinderbox using clang instead of GCC.
powerpcspe is disabled for now until clang/llvm issues with spe
have been fixed.
Reviewed by: imp
Differential Revision:https://reviews.freebsd.org/D23031
Modified:
head/Makefile
Modified: head/Makefile
==
--- head/Makefile Mon Jan 6 17:34:17 2020(r356418)
+++ head/Makefile Mon Jan 6 17:36:28 2020(r356419)
@@ -491,15 +491,13 @@ worlds: .PHONY
# honor that most of all.
#
_OBSOLETE_GCC_TARGETS=mips sparc64
-.if defined(MAKE_OBSOLETE_GCC)
-_OBSOLETE_GCC_TARGETS+=powerpc
-.endif
-TARGETS?=amd64 arm arm64 i386 riscv ${_OBSOLETE_GCC_TARGETS}
+TARGETS?=amd64 arm arm64 i386 powerpc riscv ${_OBSOLETE_GCC_TARGETS}
_UNIVERSE_TARGETS= ${TARGETS}
TARGET_ARCHES_arm?=armv6 armv7
TARGET_ARCHES_arm64?= aarch64
TARGET_ARCHES_mips?= mipsel mips mips64el mips64 mipsn32 mipselhf mipshf
mips64elhf mips64hf
-TARGET_ARCHES_powerpc?=powerpc powerpc64 powerpcspe
+# powerpcspe excluded until clang fixed
+TARGET_ARCHES_powerpc?=powerpc powerpc64
# riscv64sf excluded due to PR 232085
TARGET_ARCHES_riscv?= riscv64
.for target in ${TARGETS}
@@ -510,12 +508,10 @@ MAKE_PARAMS_riscv?= CROSS_TOOLCHAIN=riscv64-gcc
.if !defined(MAKE_OBSOLETE_GCC)
OBSOLETE_GCC_TARGETS=${_OBSOLETE_GCC_TARGETS}
MAKE_PARAMS_mips?= CROSS_TOOLCHAIN=mips-gcc6
-MAKE_PARAMS_powerpc?= CROSS_TOOLCHAIN=powerpc64-gcc6
MAKE_PARAMS_sparc64?= CROSS_TOOLCHAIN=sparc64-gcc6
.endif
TOOLCHAINS_mips= mips-gcc6
-TOOLCHAINS_powerpc=powerpc64-gcc6
TOOLCHAINS_riscv= riscv64-gcc
TOOLCHAINS_sparc64=sparc64-gcc6
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r356418 - head
Author: jhb
Date: Mon Jan 6 17:34:17 2020
New Revision: 356418
URL: https://svnweb.freebsd.org/changeset/base/356418
Log:
Only build kernels for enabled TARGET_ARCHes in make universe/tinderbox.
Previously, all of the kernels for a given TARGET were built if that
target was enabled. This was implemented by having each kernel built
via a universe_kernconf_ target that was depended on by a
universe_kernconfs target. However, this meant that if one did a
build with a limited set of TARGET_ARCH values for a given TARGET,
kernels could be built for which we hadn't built a world or toolchain.
For example, 'make TARGETS=mips TARGET_ARCHES_mips=mips64' would build
mips32 kernels.
Fix this by adding an extra layer of indirection in the kernel make
targets. universe_kernconf_ is now a dependency of a new
universe_kernconfs_. universe_kernconfs in turn depends
on a list of universe_kernconfs_ values, but only the
values enabled in TARGET_ARCHES_.
Reviewed by: imp
MFC after:1 month
Differential Revision:https://reviews.freebsd.org/D23031
Modified:
head/Makefile
Modified: head/Makefile
==
--- head/Makefile Mon Jan 6 15:29:14 2020(r356417)
+++ head/Makefile Mon Jan 6 17:34:17 2020(r356418)
@@ -724,7 +724,7 @@ TARGET_ARCH_${kernel}!= cd ${KERNSRCDIR}/${TARGET}/con
.if empty(TARGET_ARCH_${kernel})
.error "Target architecture for ${TARGET}/conf/${kernel} unknown. config(8)
likely too old."
.endif
-universe_kernconfs: universe_kernconf_${TARGET}_${kernel}
+universe_kernconfs_${TARGET_ARCH_${kernel}}:
universe_kernconf_${TARGET}_${kernel}
universe_kernconf_${TARGET}_${kernel}: .MAKE
@echo ">> ${TARGET}.${TARGET_ARCH_${kernel}} ${kernel} kernel started
on `LC_ALL=C date`"
@(cd ${.CURDIR} && env __MAKE_CONF=/dev/null \
@@ -737,6 +737,9 @@ universe_kernconf_${TARGET}_${kernel}: .MAKE
(echo "${TARGET} ${kernel} kernel failed," \
"check _.${TARGET}.${kernel} for details"| ${MAKEFAIL}))
@echo ">> ${TARGET}.${TARGET_ARCH_${kernel}} ${kernel} kernel completed
on `LC_ALL=C date`"
+.endfor
+.for target_arch in ${TARGET_ARCHES_${TARGET}}
+universe_kernconfs: universe_kernconfs_${target_arch} .PHONY
.endfor
.endif # make(universe_kernels)
universe: universe_epilogue
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
