svn commit: r252498 - head/sys/dev/usb
Author: rpaulo Date: Tue Jul 2 06:49:05 2013 New Revision: 252498 URL: http://svnweb.freebsd.org/changeset/base/252498 Log: Add RealTek 8192SU product IDs. Modified: head/sys/dev/usb/usbdevs Modified: head/sys/dev/usb/usbdevs == --- head/sys/dev/usb/usbdevsTue Jul 2 04:45:51 2013(r252497) +++ head/sys/dev/usb/usbdevsTue Jul 2 06:49:05 2013(r252498) @@ -849,6 +849,7 @@ product ACCTON RT3070_2 0xa702 RT3070 product ACCTON RT2870_10xb522 RT2870 productACCTON RT3070_3 0xc522 RT3070 productACCTON RT3070_5 0xd522 RT3070 +product ACCTON RTL8192SU0xc512 RTL8192SU product ACCTON ZD1211B 0xe501 ZD1211B /* Aceeca products */ @@ -1134,7 +1135,9 @@ product ASUS RT2870_4 0x1760 RT2870 product ASUS RT2870_5 0x1761 RT2870 productASUS USBN13 0x1784 USB-N13 productASUS RT3070_1 0x1790 RT3070 +product ASUS USBN100x1786 USB-N10 product ASUS RTL8192CU 0x17ab RTL8192CU +product ASUS RTL8192SU 0x1791 RTL8192SU product ASUS A730W 0x4202 ASUS MyPal A730W product ASUS P535 0x420f ASUS P535 PDA productASUS GMSC 0x422f ASUS Generic Mass Storage @@ -1185,7 +1188,11 @@ product AZUREWAVE RT3070_3 0x3305 RT3070 product AZUREWAVE RTL8188CU0x3357 RTL8188CU product AZUREWAVE RTL8188CE_1 0x3358 RTL8188CE product AZUREWAVE RTL8188CE_2 0x3359 RTL8188CE - +product AZUREWAVE RTL8192SU_1 0x3306 RTL8192SU +product AZUREWAVE RTL8192SU_2 0x3309 RTL8192SU +product AZUREWAVE RTL8192SU_3 0x3310 RTL8192SU +product AZUREWAVE RTL8192SU_4 0x3311 RTL8192SU +product AZUREWAVE RTL8192SU_5 0x3325 RTL8192SU /* Baltech products */ product BALTECH CARDREADER 0x Card reader @@ -1242,6 +1249,9 @@ product BELKIN F5D7050E 0x705e F5D7050E product BELKIN RT2870_10x8053 RT2870 product BELKIN RT2870_20x805c RT2870 product BELKIN F5D8053V3 0x815c F5D8053 v3 +product BELKIN RTL8192SU_1 0x815f RTL8192SU +product BELKIN RTL8192SU_2 0x845a RTL8192SU +product BELKIN RTL8192SU_3 0x945a RTL8192SU product BELKIN F5D8055 0x825a F5D8055 product BELKIN F5D8055V2 0x825b F5D8055 v2 product BELKIN F5D9050V3 0x905b F5D9050 ver 3 Wireless Adapter @@ -1368,6 +1378,9 @@ product CONCEPTRONIC AR5523_1 0x7801 AR5 product CONCEPTRONIC AR5523_1_NF 0x7802 AR5523 (no firmware) product CONCEPTRONIC AR5523_2 0x7811 AR5523 product CONCEPTRONIC AR5523_2_NF 0x7812 AR5523 (no firmware) +product CONCEPTRONIC2 RTL8192SU_1 0x3300 RTL8192SU +product CONCEPTRONIC2 RTL8192SU_2 0x3301 RTL8192SU +product CONCEPTRONIC2 RTL8192SU_3 0x3302 RTL8192SU product CONCEPTRONIC2 C54RU0x3c02 C54RU WLAN product CONCEPTRONIC2 C54RU2 0x3c22 C54RU product CONCEPTRONIC2 RT3070_1 0x3c08 RT3070 @@ -1404,6 +1417,7 @@ product COREGA RT2870_2 0x003c RT2870 product COREGA RT2870_30x003f RT2870 product COREGA RT3070 0x0041 RT3070 product COREGA CGWLUSB300GNM 0x0042 CG-WLUSB300GNM +product COREGA RTL8192SU0x0047 RTL8192SU product COREGA RTL8192CU 0x0056 RTL8192CU product COREGA WLUSB_11_STICK 0x7613 WLAN USB Stick 11 @@ -1546,6 +1560,9 @@ product DLINK RTL8192CU_1 0x3307 RTL8192 productDLINK RTL8192CU_2 0x3309 RTL8192CU productDLINK RTL8192CU_3 0x330a RTL8192CU productDLINK DWA131B 0x330d DWA-131 rev B +product DLINK2 RTL8192SU_1 0x3300 RTL8192SU +product DLINK2 RTL8192SU_2 0x3302 RTL8192SU +product DLINK2 DWA131A10x3303 DWA-131 A1 product DLINK2 DWA120 0x3a0c DWA-120 product DLINK2 DWA120_NF 0x3a0d DWA-120 (no firmware) product DLINK2 DWLG122C1 0x3c03 DWL-G122 c1 @@ -1582,6 +1599,9 @@ product DYNASTREAM ANTDEVBOARD2 0x1006 A /* Edimax products */ product EDIMAX EW7318USG 0x7318 USB Wireless dongle +product EDIMAX RTL8192SU_1 0x7611 RTL8192SU +product EDIMAX RTL8192SU_2 0x7612 RTL8192SU +product EDIMAX RTL8192SU_3 0x7622 RTL8192SU product EDIMAX RT2870_10x7711 RT2870 product EDIMAX EW7717 0x7717 EW-7717 product EDIMAX EW7718 0x7718 EW-7718 @@ -2075,6 +2095,8 @@ product GUILLEMOT HWGUSB254 0xe000 HWGUS product GUILLEMOT HWGUSB254LB 0xe010 HWGUSB2-54-LB product GUILLEMOT HWGUSB254V2AP0xe020 HWGUSB2-54V2-AP product GUILLEMOT HWNU300 0xe030 HWNU-300 +product GUILLEMOT HWNUM300 0xe031 HWNUm-300 +product GUILLEMOT HWGUN54 0xe032 HWGUn-54 product GUILLEMOT HWNUP150 0xe033 HWNUP-150 /* Hagiwara products */ @@ -2101,6 +2123,8 @@ product HAWKING HWUN2 0x0009 HWUN2 product HAWKING RT3070 0x000b
svn commit: r252499 - head/sys/powerpc/wii
Author: rpaulo
Date: Tue Jul 2 06:59:37 2013
New Revision: 252499
URL: http://svnweb.freebsd.org/changeset/base/252499
Log:
Add register definitions for the Wii IPC system.
Modified:
head/sys/powerpc/wii/wii_ipcreg.h
Modified: head/sys/powerpc/wii/wii_ipcreg.h
==
--- head/sys/powerpc/wii/wii_ipcreg.h Tue Jul 2 06:49:05 2013
(r252498)
+++ head/sys/powerpc/wii/wii_ipcreg.h Tue Jul 2 06:59:37 2013
(r252499)
@@ -34,4 +34,69 @@
#defineWIIIPC_IOH_ADDR 0x133e
#defineWIIIPC_IOH_LEN 0xc2
+#defineWIIIPC_TXBUF0x00
+#defineWIIIPC_CSR 0x04
+#defineWIIIPC_CSR_TXSTART 0x01
+#defineWIIIPC_CSR_TBEI 0x02
+#defineWIIIPC_CSR_RBFI 0x04
+#defineWIIIPC_CSR_RXREADY 0x08
+#defineWIIIPC_CSR_RBFIMASK 0x10
+#defineWIIIPC_CSR_TBEIMASK 0x20
+#defineWIIIPC_RXBUF0x08
+#defineWIIIPC_ISR 0x30
+#defineWIIIPC_ISR_MAGIC0x4000
+
+enum wiiipc_cmd {
+ WIIIPC_CMD_OPEN = 1,
+ WIIIPC_CMD_CLOSE= 2,
+ WIIIPC_CMD_READ = 3,
+ WIIIPC_CMD_WRITE= 4,
+ WIIIPC_CMD_SEEK = 5,
+ WIIIPC_CMD_IOCTL= 6,
+ WIIIPC_CMD_IOCTLV = 7,
+ WIIIPC_CMD_ASYNCRESP= 8
+};
+
+struct wiiipc_ipc_msg {
+ uint32_tipc_cmd;
+ int32_t ipc_result;
+ int32_t ipc_fd; /* WIIIPC_CMD_ASYNCRESP - the original cmd */
+ union {
+ struct {
+ intptr_t pathname;
+ uint32_t mode;
+ } _ipc_open;
+ struct {
+ intptr_t data;
+ uint32_t len;
+ } _ipc_read, _ipc_write;
+ struct {
+ int32_t offset;
+ int32_t whence;
+ } _ipc_seek;
+ struct {
+ uint32_t request;
+ intptr_t ibuf;
+ uint32_t ilen;
+ intptr_t obuf;
+ uint32_t olen;
+ } _ipc_ioctl;
+ struct {
+ uint32_t request;
+ uint32_t argin;
+ uint32_t argout;
+ intptr_t iovec;
+ } _ipc_ioctlv;
+ uint32_t _ipc_argv[5];
+ } args;
+} __attribute__((packed));
+
+CTASSERT(sizeof(struct wiiipc_ipc_msg) == 32);
+
+#define ipc_open args._ipc_open
+#defineipc_readargs._ipc_read
+#defineipc_write args._ipc_write
+#define ipc_ioctl args._ipc_ioctl
+#define ipc_ioctlv args._ipc_ioctlv
+
#endif /* _POWERPC_WII_WII_IPCREG_H */
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r252500 - head/sys/powerpc/wii
Author: rpaulo
Date: Tue Jul 2 07:17:56 2013
New Revision: 252500
URL: http://svnweb.freebsd.org/changeset/base/252500
Log:
Fix indentation.
Submitted by: jmallet
Modified:
head/sys/powerpc/wii/wii_ipcreg.h
Modified: head/sys/powerpc/wii/wii_ipcreg.h
==
--- head/sys/powerpc/wii/wii_ipcreg.h Tue Jul 2 06:59:37 2013
(r252499)
+++ head/sys/powerpc/wii/wii_ipcreg.h Tue Jul 2 07:17:56 2013
(r252500)
@@ -93,10 +93,10 @@ struct wiiipc_ipc_msg {
CTASSERT(sizeof(struct wiiipc_ipc_msg) == 32);
-#define ipc_open args._ipc_open
+#defineipc_openargs._ipc_open
#defineipc_readargs._ipc_read
#defineipc_write args._ipc_write
-#define ipc_ioctl args._ipc_ioctl
-#define ipc_ioctlv args._ipc_ioctlv
+#defineipc_ioctl args._ipc_ioctl
+#defineipc_ioctlv args._ipc_ioctlv
#endif /* _POWERPC_WII_WII_IPCREG_H */
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r252501 - in head/sys/dev/usb: . quirk
Author: hrs Date: Tue Jul 2 07:35:57 2013 New Revision: 252501 URL: http://svnweb.freebsd.org/changeset/base/252501 Log: Add Planex MZK-UE150N. Submitted by: Yusuke Tanaka MFC after:3 days Modified: head/sys/dev/usb/quirk/usb_quirk.c head/sys/dev/usb/usbdevs Modified: head/sys/dev/usb/quirk/usb_quirk.c == --- head/sys/dev/usb/quirk/usb_quirk.c Tue Jul 2 07:17:56 2013 (r252500) +++ head/sys/dev/usb/quirk/usb_quirk.c Tue Jul 2 07:35:57 2013 (r252501) @@ -94,6 +94,7 @@ static struct usb_quirk_entry usb_quirks USB_QUIRK(SILICONPORTALS, YAPPHONE, 0x100, 0x100, UQ_AU_INP_ASYNC), USB_QUIRK(LOGITECH, UN53B, 0x, 0x, UQ_NO_STRINGS), USB_QUIRK(ELSA, MODEM1, 0x, 0x, UQ_CFG_INDEX_1), + USB_QUIRK(PLANEX2, MZKUE150N, 0x, 0x, UQ_CFG_INDEX_1), /* Quirks for printer devices */ USB_QUIRK(HP, 895C, 0x, 0x, UQ_BROKEN_BIDIR), USB_QUIRK(HP, 880C, 0x, 0x, UQ_BROKEN_BIDIR), Modified: head/sys/dev/usb/usbdevs == --- head/sys/dev/usb/usbdevsTue Jul 2 07:17:56 2013(r252500) +++ head/sys/dev/usb/usbdevsTue Jul 2 07:35:57 2013(r252501) @@ -3351,6 +3351,7 @@ product PLANEX2 RTL8192CU 0xab2b RTL8192 product PLANEX2 GWUS54HP 0xab01 GW-US54HP product PLANEX2 GWUS300MINIS 0xab24 GW-US300MiniS product PLANEX2RT3070 0xab25 RT3070 +product PLANEX2 MZKUE150N 0xab2f MZK-UE150N product PLANEX2 GWUS54MINI20xab50 GW-US54Mini2 product PLANEX2 GWUS54SG 0xc002 GW-US54SG product PLANEX2 GWUS54GZL 0xc007 GW-US54GZL ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r252502 - head/sys/kern
Author: mjg
Date: Tue Jul 2 07:36:04 2013
New Revision: 252502
URL: http://svnweb.freebsd.org/changeset/base/252502
Log:
Fix receiving fd over unix socket broken in r247740.
If n fds were passed, it would receive the first one n times.
Reported by: Shawn Webb latt...@gmail.com, koobs, gleb
Tested by:koobs, gleb
Reviewed by: pjd
Modified:
head/sys/kern/uipc_usrreq.c
Modified: head/sys/kern/uipc_usrreq.c
==
--- head/sys/kern/uipc_usrreq.c Tue Jul 2 07:35:57 2013(r252501)
+++ head/sys/kern/uipc_usrreq.c Tue Jul 2 07:36:04 2013(r252502)
@@ -1764,8 +1764,8 @@ unp_externalize(struct mbuf *control, st
}
for (i = 0; i newfds; i++, fdp++) {
fde = fdesc-fd_ofiles[*fdp];
- fde-fde_file = fdep[0]-fde_file;
- filecaps_move(fdep[0]-fde_caps,
+ fde-fde_file = fdep[i]-fde_file;
+ filecaps_move(fdep[i]-fde_caps,
fde-fde_caps);
if ((flags MSG_CMSG_CLOEXEC) != 0)
fde-fde_flags |= UF_EXCLOSE;
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
RAND_MAX broken (was: Re: svn commit: r252484 - head/sys/ufs/ffs)
On Tue, 2 Jul 2013, Bruce Evans wrote:
...
Also, random(9) is internally broken on 64-bit arches. It shouldn't
exist. It is not random() at all, but just rand() with the clamp to
RAND_MAX removed. Its linear congruential generator is suboptimal
for 64 bits, and other parts of its algorithm are just broken for
64 bits. It uses longs internally to try to fake unsigned 31 bits,
but the faking is broken in different ways depending on the size of
long. It is documented to return a 31-bit value, but on 32-bit
arches it seems to be possible for it to return 0xf, and on
64-bit arches it does return full 64-bit values.
random(9) was cloned from rand(3). The userland versions have been
edited a bit, but still have most of the bugs:
random(3) uses an internal copy of rand(3) named good_rand() for
seeding. If USE_WEAK_SEEDING is defined, this uses a weaker linear
congruential generator (LCG). This uses int32_t instead of long
internally, and doesn't attempt to reduce the value to 31 bits.
Otherwise, the same LCG as random(9) is used, with the same buggy code
except for changing the longs to 32-bits. Using int32_t gives the
same overflow bugs on all (supported) arches, and avoids returning
invalid values half the time on 64-bit arches. There are 2 callers
of good_rand(), and only 1 of them discards bits above the 31st.
srand() also supports USE_WEAK_SEEDING. It uses u_long internally,
so it is actually correct. The internal value has the number of
bits in a u_long and is generated without overflow and without any
bias in the reduction to 31 bits. Then returning this value as
an int in by taking the value modulo ((u_long)RAND_MAX + 1) gives
a correct reduction to 31 bits when RAND_MAX is 0x7fff (or
15 bits if RAND_MAX is 0x7fff, etc.).
srand() in the !USE_WEAK_SEEDING case still uses the same buggy code
as random(9), with type long, so the internal values overflow and the
inernal reduction to 31 bits is buggy, with the bugs depending on the
size of long and other things. But it is mostly saved by taking the
value modulo ((u_long)RAND_MAX + 1). This reduces to a valid value
and leaves only minor biases from the buggy earlier reduction.
random(6) used to have bugs related to the buggy internal reduction,
and the biases from these were noticed. It uses floating point, so
the reduction was easier, but it was still done wrong, by dividing
by LONG_MAX instead of RANDOM_MAX_PLUS1. Using LONG_MAX is like using
0x7fff in random(9), but obviously buggier since the range for
both is documented to be [0,0x7fff]. Hard-coding 0x7fff in
random(6) would have been equally buggy. I think 0x8000 is correct
in both, but in random(9) this assumes too much about type sizes and
layouts. The correct method in integer code is to take an unsigned
modulus with divisor 0x8000. Let the compiler optimize this to
masking with 0x7fff. This depends on the maximum value plus 1
being representable in an unsigned type. For rand(), this occurs
because rand() returns int, and for random(3), this occurs because
random(3) returns long.
Another bug in random(9) is that it returns u_long, so its API is
different from random(3), but since it is documented to return only
31 bits, it is not really different except in the buggy cases where
it returns 32-64 bits.
The bugs are a little different than I said above. There is no overflow
problem and no problem with invalid values being produces, since the
algorithm from ACM is careful to do everything with 32 bit signed
integers without causing overflow. The algorithm just doesn't produce
values mod 2**32 as expected by all the functions. It does what it
claims to do -- it produces values mod (2**32 - 1). The largest bug
is that RAND_MAX is off by 1. It is specified as the largest value
returned by rand(), but in FreeBSD rand() never returns it unless
USE_WEAK_SEEDING is confgured. The values should be unifornly
distributed on average beteen 0 and RAND_MAX,but that is impossible
if RADND_MAX is never returned. From libc/stdlib/srand.c:
% static int
% do_rand(unsigned long *ctx)
% {
% #ifdef USE_WEAK_SEEDING
% /*
% * Historic implementation compatibility.
% * The random sequences do not vary much with the seed,
% * even with overflowing.
% */
% return ((*ctx = *ctx * 1103515245 + 12345) % ((u_long)RAND_MAX + 1));
RAND_MAX is correct for this algorithm. Except it is off by about a factor
of 65536 instead of just by +1. This is the LCG documented as an example in
C standards. It is only suitable for use with 16-bit ints and/or with
RAND_MAX = 32767 and that is how it is used in the example. With FreeBSD's
RAND_MAX = 0x7fff, RAND_MAX is correct but the implementation is of low
quality.
% #else /* !USE_WEAK_SEEDING */
% /*
% * Compute x = (7^5 * x) mod (2^31 - 1)
Note that (2**31 - 1) is not 2**31.
% * without overflowing 31 bits:
% * (2^31 - 1) = 127773 * (7^5) + 2836
% * From Random number generators: good ones
svn commit: r252503 - head/contrib/llvm/lib/Target/ARM
Author: andrew
Date: Tue Jul 2 08:04:41 2013
New Revision: 252503
URL: http://svnweb.freebsd.org/changeset/base/252503
Log:
Work around an ARM EABI issue where clang would sometimes incorrectly align
the stack in a leaf function that uses TLS.
The issue is, when using TLS, the function is no longer a leaf as it calls
__aeabi_read_tp. With statically linked programs this is not an issue as
it doesn't make use of the stack, however with dynamically linked
applications we enter rtld which does use the stack and makes assumptions
about it's alignment.
This is only a temporary fix until a better patch can be made and submitted
upstream.
Modified:
head/contrib/llvm/lib/Target/ARM/ARMFrameLowering.h
Modified: head/contrib/llvm/lib/Target/ARM/ARMFrameLowering.h
==
--- head/contrib/llvm/lib/Target/ARM/ARMFrameLowering.h Tue Jul 2 07:36:04
2013(r252502)
+++ head/contrib/llvm/lib/Target/ARM/ARMFrameLowering.h Tue Jul 2 08:04:41
2013(r252503)
@@ -27,7 +27,7 @@ protected:
public:
explicit ARMFrameLowering(const ARMSubtarget sti)
-: TargetFrameLowering(StackGrowsDown, sti.getStackAlignment(), 0, 4),
+: TargetFrameLowering(StackGrowsDown, sti.getStackAlignment(), 0, 8),
STI(sti) {
}
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r252504 - in head: share/man/man4 sys/modules/cc sys/modules/cc/cc_cdg sys/netinet/cc
Author: lstewart Date: Tue Jul 2 08:44:56 2013 New Revision: 252504 URL: http://svnweb.freebsd.org/changeset/base/252504 Log: Import an implementation of the CAIA Delay-Gradient (CDG) congestion control algorithm, which is based on the 2011 v0.1 patch release and described in the paper Revisiting TCP Congestion Control using Delay Gradients by David Hayes and Grenville Armitage. It is implemented as a kernel module compatible with the modular congestion control framework. CDG is a hybrid congestion control algorithm which reacts to both packet loss and inferred queuing delay. It attempts to operate as a delay-based algorithm where possible, but utilises heuristics to detect loss-based TCP cross traffic and will compete effectively as required. CDG is therefore incrementally deployable and suitable for use on shared networks. In collaboration with:David Hayes david.hayes at ieee.org and Grenville Armitage garmitage at swin edu au MFC after:4 days Sponsored by: Cisco University Research Program and FreeBSD Foundation Added: head/share/man/man4/cc_cdg.4 (contents, props changed) head/sys/modules/cc/cc_cdg/ head/sys/modules/cc/cc_cdg/Makefile (contents, props changed) head/sys/netinet/cc/cc_cdg.c (contents, props changed) Modified: head/share/man/man4/Makefile head/sys/modules/cc/Makefile Modified: head/share/man/man4/Makefile == --- head/share/man/man4/MakefileTue Jul 2 08:04:41 2013 (r252503) +++ head/share/man/man4/MakefileTue Jul 2 08:44:56 2013 (r252504) @@ -81,6 +81,7 @@ MAN= aac.4 \ cardbus.4 \ carp.4 \ cas.4 \ + cc_cdg.4 \ cc_chd.4 \ cc_cubic.4 \ cc_hd.4 \ Added: head/share/man/man4/cc_cdg.4 == --- /dev/null 00:00:00 1970 (empty, because file is newly added) +++ head/share/man/man4/cc_cdg.4Tue Jul 2 08:44:56 2013 (r252504) @@ -0,0 +1,155 @@ +.\ +.\ Copyright (c) 2013 Swinburne University of Technology, Melbourne, Australia +.\ All rights reserved. +.\ +.\ Redistribution and use in source and binary forms, with or without +.\ modification, are permitted provided that the following conditions +.\ are met: +.\ 1. Redistributions of source code must retain the above copyright +.\notice, this list of conditions and the following disclaimer. +.\ 2. Redistributions in binary form must reproduce the above copyright +.\notice, this list of conditions and the following disclaimer in the +.\documentation and/or other materials provided with the distribution. +.\ +.\ THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND +.\ ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE +.\ IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE +.\ ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE FOR +.\ ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL +.\ DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS +.\ OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) +.\ HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT +.\ LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY +.\ OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF +.\ SUCH DAMAGE. +.\ +.\ $FreeBSD$ +.\ +.Dd July 2, 2013 +.Dt CC_CDG 4 +.Os +.Sh NAME +.Nm cc_cdg +.Nd CDG Congestion Control Algorithm +.Sh DESCRIPTION +CAIA-Delay Gradient (CDG) is a hybrid congestion control algorithm which reacts +to both packet loss and inferred queuing delay. +It attempts to operate as a delay-based algorithm where possible, but utilises +heuristics to detect loss-based TCP cross traffic and will compete effectively +as required. +CDG is therefore incrementally deployable and suitable for use on shared +networks. +.Pp +During delay-based operation, CDG uses a delay-gradient based probabilistic +backoff mechanism, and will also try to infer non congestion related +packet losses and avoid backing off when they occur. +During loss-based operation, CDG essentially reverts to +.Xr cc_newreno 4 Ns - Ns like +behaviour. +.Pp +CDG switches to loss-based operation when it detects that a configurable number +of consecutive delay-based backoffs have had no measurable effect. +It periodically attempts to return to delay-based operation, but will keep +switching back to loss-based operation as required. +.Sh MIB Variables +The algorithm exposes the following variables in the +.Va net.inet.tcp.cc.cdg +branch of the +.Xr sysctl 3 +MIB: +.Bl -tag -width .Va exp_backoff_scale +.It Va version +Current algorithm/implementation version number. +.It Va beta_delay +Delay-based window decrease factor as a percentage (on delay-based backoff, w = +w *
svn commit: r252505 - head/sys/conf
Author: gjb
Date: Tue Jul 2 10:36:57 2013
New Revision: 252505
URL: http://svnweb.freebsd.org/changeset/base/252505
Log:
- Update newvers.sh to include svn revision in uname(1) if the
system has svnliteversion.
- If svnliteversion is not found, look for svnversion in /usr/bin
and /usr/local/bin, since svnlite can be installed as svn if
WITH_SVN is set.[1]
- Remove /bin from binary search paths.[1]
Discussed with: kib [1]
MFC after:3 days
Approved by: kib (mentor)
Modified:
head/sys/conf/newvers.sh
Modified: head/sys/conf/newvers.sh
==
--- head/sys/conf/newvers.shTue Jul 2 08:44:56 2013(r252504)
+++ head/sys/conf/newvers.shTue Jul 2 10:36:57 2013(r252505)
@@ -88,16 +88,26 @@ v=`cat version` u=${USER:-root} d=`pwd`
i=`${MAKE:-make} -V KERN_IDENT`
compiler_v=$($(${MAKE:-make} -V CC) -v 21 | grep 'version')
-for dir in /bin /usr/bin /usr/local/bin; do
+if [ -x /usr/bin/svnliteversion ] ; then
+ svnversion=/usr/bin/svnliteversion
+fi
+
+for dir in /usr/bin /usr/local/bin; do
+ if [ ! -z ${svnversion} ] ; then
+ break
+ fi
if [ -x ${dir}/svnversion ] [ -z ${svnversion} ] ; then
svnversion=${dir}/svnversion
+ break
fi
+done
+for dir in /usr/bin /usr/local/bin; do
if [ -x ${dir}/p4 ] [ -z ${p4_cmd} ] ; then
p4_cmd=${dir}/p4
fi
done
if [ -d ${SYSDIR}/../.git ] ; then
- for dir in /bin /usr/bin /usr/local/bin; do
+ for dir in /usr/bin /usr/local/bin; do
if [ -x ${dir}/git ] ; then
git_cmd=${dir}/git --git-dir=${SYSDIR}/../.git
break
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r252506 - head/sbin/dhclient
Author: bms
Date: Tue Jul 2 13:24:37 2013
New Revision: 252506
URL: http://svnweb.freebsd.org/changeset/base/252506
Log:
When acquiring a lease, record the value of the BOOTP siaddr field
contained in the DHCP offer, and write it out to the lease file
as an unquoted value of the next-server keyword. The value is ignored
when the lease is read back by dhclient, however other applications
are free to parse it.
The intent behind this change is to allow easier interoperability
with automated installation systems e.g. Cobbler, Foreman, Razor;
FreeBSD installation kernels can automatically probe the network
to discover deployment servers. There are no plans to MFC this
change unless a backport is specifically requested.
The syntax of the next-server ip lease keyword is intended to be
identical to that used by the ISC DHCPD server in its configuration files.
The required defines are already present in dhclient but were unused before
this change. (Note: This is NOT the same as Option 66, tftp-server-name).
It has been exercised in a university protocol testbed environment, with
Cobbler and an mfsBSD image containing pc-sysinstall (driven by Cobbler
Cheetah templates). The SYSLINUX memdisk driver is used to boot mfsBSD.
Currently this approach requires that a dedicated system profile has
been created for the node where FreeBSD is to be deployed. If this
is not present, the pc-sysinstall wrapper will be unable to obtain
a node configuration. There is code in progress to allow mfsBSD images
to obtain the required hints from the memdisk environment by parsing
the MBFT ACPI chunk. This is non-standard as it is not linked into
the platform's ACPI RSDT.
Reviewed by: des
Modified:
head/sbin/dhclient/clparse.c
head/sbin/dhclient/dhclient.c
head/sbin/dhclient/dhcpd.h
Modified: head/sbin/dhclient/clparse.c
==
--- head/sbin/dhclient/clparse.cTue Jul 2 10:36:57 2013
(r252505)
+++ head/sbin/dhclient/clparse.cTue Jul 2 13:24:37 2013
(r252506)
@@ -642,6 +642,10 @@ parse_client_lease_declaration(FILE *cfi
case FILENAME:
lease-filename = parse_string(cfile);
return;
+ case NEXT_SERVER:
+ if (!parse_ip_addr(cfile, lease-nextserver))
+ return;
+ break;
case SERVER_NAME:
lease-server_name = parse_string(cfile);
return;
Modified: head/sbin/dhclient/dhclient.c
==
--- head/sbin/dhclient/dhclient.c Tue Jul 2 10:36:57 2013
(r252505)
+++ head/sbin/dhclient/dhclient.c Tue Jul 2 13:24:37 2013
(r252506)
@@ -1063,6 +1063,9 @@ packet_to_lease(struct packet *packet)
lease-address.len = sizeof(packet-raw-yiaddr);
memcpy(lease-address.iabuf, packet-raw-yiaddr, lease-address.len);
+ lease-nextserver.len = sizeof(packet-raw-siaddr);
+ memcpy(lease-nextserver.iabuf, packet-raw-siaddr,
lease-nextserver.len);
+
/* If the server name was filled out, copy it.
Do not attempt to validate the server name as a host name.
RFC 2131 merely states that sname is NUL-terminated (which do
@@ -1874,6 +1877,11 @@ write_client_lease(struct interface_info
fprintf(leaseFile, bootp;\n);
fprintf(leaseFile, interface \%s\;\n, ip-name);
fprintf(leaseFile, fixed-address %s;\n, piaddr(lease-address));
+ if (lease-nextserver.len == sizeof(inaddr_any)
+ 0 != memcmp(lease-nextserver.iabuf, inaddr_any,
+ sizeof(inaddr_any)))
+ fprintf(leaseFile, next-server %s;\n,
+ piaddr(lease-nextserver));
if (lease-filename)
fprintf(leaseFile, filename \%s\;\n, lease-filename);
if (lease-server_name)
Modified: head/sbin/dhclient/dhcpd.h
==
--- head/sbin/dhclient/dhcpd.h Tue Jul 2 10:36:57 2013(r252505)
+++ head/sbin/dhclient/dhcpd.h Tue Jul 2 13:24:37 2013(r252506)
@@ -121,6 +121,7 @@ struct client_lease {
struct client_lease *next;
time_t expiry, renewal, rebind;
struct iaddr address;
+ struct iaddr nextserver;
char*server_name;
char*filename;
struct string_list *medium;
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r252508 - head/sbin/devd
Author: asomers Date: Tue Jul 2 14:48:39 2013 New Revision: 252508 URL: http://svnweb.freebsd.org/changeset/base/252508 Log: Explicitly include cstdarg to fix compilation with libc++. It is implicitly included by libstdc++. Reported By: Oliver Hartmann Approved by: gibbs (co-mentor, implicit) Modified: head/sbin/devd/devd.cc Modified: head/sbin/devd/devd.cc == --- head/sbin/devd/devd.cc Tue Jul 2 13:48:16 2013(r252507) +++ head/sbin/devd/devd.cc Tue Jul 2 14:48:39 2013(r252508) @@ -79,6 +79,7 @@ __FBSDID($FreeBSD$); #include cstdio #include csignal #include cstring +#include cstdarg #include dirent.h #include err.h ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: RAND_MAX broken
On 02.07.2013 11:39, Bruce Evans wrote: The bugs are a little different than I said above. There is no overflow problem and no problem with invalid values being produces, since the algorithm from ACM is careful to do everything with 32 bit signed integers without causing overflow. The algorithm just doesn't produce values mod 2**32 as expected by all the functions. It does what it claims to do -- it produces values mod (2**32 - 1). The largest bug is that RAND_MAX is off by 1. It is specified as the largest value returned by rand(), but in FreeBSD rand() never returns it unless USE_WEAK_SEEDING is confgured. The values should be unifornly distributed on average beteen 0 and RAND_MAX,but that is impossible if RADND_MAX is never returned. From libc/stdlib/srand.c: Don't ever consider USE_WEAK_SEEDING defined - result is distributet _very_ poorly and the code should be removed long time ago. BTW, I don't understand well fixes you suggest. Is it to define RAND_MAX as 0x7ffe ? -- http://ache.vniz.net/ bitcoin:13fGiNutKNHcVSsgtGQ7bQ5kgUKgEQHn7N ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: svn commit: r252346 - head/share/man/man9
On 6/29/13 6:37 AM, Julian Elischer wrote: On 6/29/13 9:16 PM, Julian Elischer wrote: thanks! -If the owner is not currently actually running then the spin step is skipped. +then a thread attempting to acquire the mutex will spin rather than yielding +the processor. Am I wrong in thinking that it will only spin for a short while, eventually yielding? The original text said this but the new text implies it will spin forever. It does not use a spin timeout. I considered adding a separate note to define adaptive spinning generically as it is used by multiple lock types (mutexes, rwlocks, and sx locks). -- John Baldwin ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: RAND_MAX broken
On Tue, 2 Jul 2013, Andrey Chernov wrote: On 02.07.2013 11:39, Bruce Evans wrote: The bugs are a little different than I said above. There is no overflow problem and no problem with invalid values being produces, since the algorithm from ACM is careful to do everything with 32 bit signed integers without causing overflow. The algorithm just doesn't produce values mod 2**32 as expected by all the functions. It does what it claims to do -- it produces values mod (2**32 - 1). The largest bug is that RAND_MAX is off by 1. It is specified as the largest value returned by rand(), but in FreeBSD rand() never returns it unless USE_WEAK_SEEDING is confgured. The values should be unifornly distributed on average beteen 0 and RAND_MAX,but that is impossible if RADND_MAX is never returned. From libc/stdlib/srand.c: Don't ever consider USE_WEAK_SEEDING defined - result is distributet _very_ poorly and the code should be removed long time ago. BTW, I don't understand well fixes you suggest. Is it to define RAND_MAX as 0x7ffe ? That would would be more correct. It might have compatibility problems. I checked the values returned by rand(). The ACM part works as intended, so it never returns RAND_MAX. It also never returns 0. So the distribution of values in the documented range [0, RAND_MAX] is very non-uniform. It is uniform in [1, RAND_MAX - 1]. To use this algorithm for rand(), 1 should have been subtracted, giving a range of [0, 0x7ffe]. In mathematical terms, the algorithm uses the (multiplicative) group of units in the field Z/(0x7fff.Z), but for rand() we want the values to be in the the additive group Z/(0x7ffe.Z). Subtracting 1 doesn't preserve the group operation but gives the right set of values. 0x7 is used because it is prime. 0x80001, 0x and 0x10001 are not prime, so the algorithm can't easily be modified to give the range [0, 0x7fff] or a 32-bit range. The most interesting prime near 2**31 or 2*32 is 2**32 - 5. Bruce ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r252510 - head/sys/netinet
Author: hrs Date: Tue Jul 2 16:39:12 2013 New Revision: 252510 URL: http://svnweb.freebsd.org/changeset/base/252510 Log: Fix a panic when leaving MC group in a kernel with VIMAGE enabled. in_leavegroup() is called from an asynchronous task, and igmp_change_state() requires that curvnet is set by the caller. Modified: head/sys/netinet/in_mcast.c Modified: head/sys/netinet/in_mcast.c == --- head/sys/netinet/in_mcast.c Tue Jul 2 14:54:18 2013(r252509) +++ head/sys/netinet/in_mcast.c Tue Jul 2 16:39:12 2013(r252510) @@ -1236,7 +1236,9 @@ in_leavegroup_locked(struct in_multi *in KASSERT(error == 0, (%s: failed to merge inm state, __func__)); CTR1(KTR_IGMPV3, %s: doing igmp downcall, __func__); + CURVNET_SET(inm-inm_ifp-if_vnet); error = igmp_change_state(inm); + CURVNET_RESTORE(); if (error) CTR1(KTR_IGMPV3, %s: failed igmp downcall, __func__); ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
should_yield problem [Was: svn commit: r251322 - head/sys/kern]
on 03/06/2013 20:36 Konstantin Belousov said the following:
Author: kib
Date: Mon Jun 3 17:36:43 2013
New Revision: 251322
URL: http://svnweb.freebsd.org/changeset/base/251322
Log:
Be more generous when donating the current thread time to the owner of
the vnode lock while iterating over the free vnode list. Instead of
yielding, pause for 1 tick. The change is reported to help in some
virtualized environments.
Kostik,
I've just run into a problem where word generous does not seem to be
applicable at all unless I am mistaken about how the code works.
There is a thread in vdropl() call that has a vnode interlock and want to take
vnode_free_list_mtx.
Then there is a thread that keeps cycling on the locked interlock while holding
vnode_free_list_mtx:
#0 cpustop_handler () at /usr/src/sys/amd64/amd64/mp_machdep.c:1391
#1 0x80cc8e1d in ipi_nmi_handler () at
/usr/src/sys/amd64/amd64/mp_machdep.c:1373
#2 0x80cd6269 in trap (frame=0xff80002baf30) at
/usr/src/sys/amd64/amd64/trap.c:211
#3 0x80cbf86f in nmi_calltrap () at
/usr/src/sys/amd64/amd64/exception.S:501
#4 0x80924f22 in _mtx_trylock (m=0xfe00236f4c98, opts=0,
file=value optimized out, line=0) at atomic.h:160
#5 0x809d3e5a in mnt_vnode_next_active (mvp=0xff86c95e4958,
mp=0xfe00118039a8) at /usr/src/sys/kern/vfs_subr.c:4813
#6 0x809daa86 in vfs_msync (mp=0xfe00118039a8, flags=2) at
/usr/src/sys/kern/vfs_subr.c:3542
#7 0x809e06ef in sys_sync (td=value optimized out, uap=value
optimized out) at /usr/src/sys/kern/vfs_syscalls.c:149
#8 0x80cd515a in amd64_syscall (td=0xfe019d966490, traced=0) at
subr_syscall.c:135
#9 0x80cbf717 in Xfast_syscall () at
/usr/src/sys/amd64/amd64/exception.S:387
Now the curious part:
(kgdb) p td-td_swvoltick
$4 = 0
(kgdb) p td-td_ru.ru_nvcsw
$7 = 0
ticks happen to be in negative territory:
(kgdb) p ticks
$2 = -1946084020
Give this definition:
int
should_yield(void)
{
return (ticks - curthread-td_swvoltick = hogticks);
}
We see that should_yield() is going to return false until ticks roll over to
become positive again. And obviously with the thread spinning on VI_TRYLOCK()
it's not going to get its td_swvoltick updated.
I am not sure if the originally reported problem was also caused by
should_yield() or if it was something else. But in either case I think that we
should fix should_yield. Perhaps (ticks - curthread-td_swvoltick) should be
cast to unsigned before comparison?
Submitted by: Roger Pau Monn? roger@citrix.com
Discussed with: jilles
Tested by: pho
MFC after: 2 weeks
Modified:
head/sys/kern/vfs_subr.c
Modified: head/sys/kern/vfs_subr.c
==
--- head/sys/kern/vfs_subr.c Mon Jun 3 17:36:26 2013(r251321)
+++ head/sys/kern/vfs_subr.c Mon Jun 3 17:36:43 2013(r251322)
@@ -4693,7 +4693,7 @@ restart:
if (mp_ncpus == 1 || should_yield()) {
TAILQ_INSERT_BEFORE(vp, *mvp, v_actfreelist);
mtx_unlock(vnode_free_list_mtx);
- kern_yield(PRI_USER);
+ pause(vnacti, 1);
mtx_lock(vnode_free_list_mtx);
goto restart;
}
--
Andriy Gapon
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r252511 - in head/sys: net netinet6
Author: hrs
Date: Tue Jul 2 16:58:15 2013
New Revision: 252511
URL: http://svnweb.freebsd.org/changeset/base/252511
Log:
- Allow ND6_IFF_AUTO_LINKLOCAL for IFT_BRIDGE. An interface with IFT_BRIDGE
is initialized with !ND6_IFF_AUTO_LINKLOCAL !ND6_IFF_ACCEPT_RTADV
regardless of net.inet6.ip6.accept_rtadv and net.inet6.ip6.auto_linklocal.
To configure an autoconfigured link-local address (RFC 4862), the
following rc.conf(5) configuration can be used:
ifconfig_bridge0_ipv6=inet6 auto_linklocal
- if_bridge(4) now removes IPv6 addresses on a member interface to be
added when the parent interface or one of the existing member
interfaces has an IPv6 address. if_bridge(4) merges each link-local
scope zone which the member interfaces form respectively, so it causes
address scope violation. Removal of the IPv6 addresses prevents it.
- if_lagg(4) now removes IPv6 addresses on a member interfaces
unconditionally.
- Set reasonable flags to non-IPv6-capable interfaces. [*]
Submitted by: rpaulo [*]
MFC after:1 week
Modified:
head/sys/net/if_bridge.c
head/sys/net/if_lagg.c
head/sys/netinet6/in6.c
head/sys/netinet6/in6_ifattach.c
head/sys/netinet6/in6_var.h
head/sys/netinet6/nd6.c
Modified: head/sys/net/if_bridge.c
==
--- head/sys/net/if_bridge.cTue Jul 2 16:39:12 2013(r252510)
+++ head/sys/net/if_bridge.cTue Jul 2 16:58:15 2013(r252511)
@@ -118,6 +118,7 @@ __FBSDID($FreeBSD$);
#ifdef INET6
#include netinet/ip6.h
#include netinet6/ip6_var.h
+#include netinet6/in6_ifattach.h
#endif
#if defined(INET) || defined(INET6)
#include netinet/ip_carp.h
@@ -1041,14 +1042,6 @@ bridge_ioctl_add(struct bridge_softc *sc
if (ifs-if_bridge != NULL)
return (EBUSY);
- bif = malloc(sizeof(*bif), M_DEVBUF, M_NOWAIT|M_ZERO);
- if (bif == NULL)
- return (ENOMEM);
-
- bif-bif_ifp = ifs;
- bif-bif_flags = IFBIF_LEARNING | IFBIF_DISCOVER;
- bif-bif_savedcaps = ifs-if_capenable;
-
switch (ifs-if_type) {
case IFT_ETHER:
case IFT_L2VLAN:
@@ -1056,20 +1049,94 @@ bridge_ioctl_add(struct bridge_softc *sc
/* permitted interface types */
break;
default:
- error = EINVAL;
- goto out;
+ return (EINVAL);
}
+#ifdef INET6
+ /*
+* Two valid inet6 addresses with link-local scope must not be
+* on the parent interface and the member interfaces at the
+* same time. This restriction is needed to prevent violation
+* of link-local scope zone. Attempts to add a member
+* interface which has inet6 addresses when the parent has
+* inet6 triggers removal of all inet6 addresses on the member
+* interface.
+*/
+
+ /* Check if the parent interface has a link-local scope addr. */
+ if (in6ifa_llaonifp(sc-sc_ifp) != NULL) {
+ /*
+* If any, remove all inet6 addresses from the member
+* interfaces.
+*/
+ BRIDGE_XLOCK(sc);
+ LIST_FOREACH(bif, sc-sc_iflist, bif_next) {
+ if (in6ifa_llaonifp(bif-bif_ifp)) {
+ BRIDGE_UNLOCK(sc);
+ in6_ifdetach(bif-bif_ifp);
+ BRIDGE_LOCK(sc);
+ if_printf(sc-sc_ifp,
+ IPv6 addresses on %s have been removed
+ before adding it as a member to prevent
+ IPv6 address scope violation.\n,
+ bif-bif_ifp-if_xname);
+ }
+ }
+ BRIDGE_XDROP(sc);
+ if (in6ifa_llaonifp(ifs)) {
+ BRIDGE_UNLOCK(sc);
+ in6_ifdetach(ifs);
+ BRIDGE_LOCK(sc);
+ if_printf(sc-sc_ifp,
+ IPv6 addresses on %s have been removed
+ before adding it as a member to prevent
+ IPv6 address scope violation.\n,
+ ifs-if_xname);
+ }
+ } else {
+ struct in6_ifaddr *ia6_m, *ia6_s;
+ /*
+* If not, check whether one of the existing member
+* interfaces have inet6 address. If any, remove
+* inet6 addresses on the interface to be added.
+*/
+ BRIDGE_XLOCK(sc);
+ LIST_FOREACH(bif, sc-sc_iflist, bif_next) {
+ ia6_m = in6ifa_llaonifp(bif-bif_ifp);
+ if (ia6_m != NULL)
+ break;
+ }
+
svn commit: r252512 - head/gnu/usr.bin/patch
Author: obrien
Date: Tue Jul 2 17:09:57 2013
New Revision: 252512
URL: http://svnweb.freebsd.org/changeset/base/252512
Log:
Make it so that 'patch FUBAR' and 'patch -i FUBAR' operate the same.
The former makes a copy of stdin, but was not accurately putting the
content of stdin into a temp file. This lead to the undercounting
the number of lines in hunks containing NUL characters when reading
from stdin. Thus resulting in unexpected end of file in patch errors.
Modified:
head/gnu/usr.bin/patch/pch.c
Modified: head/gnu/usr.bin/patch/pch.c
==
--- head/gnu/usr.bin/patch/pch.cTue Jul 2 16:58:15 2013
(r252511)
+++ head/gnu/usr.bin/patch/pch.cTue Jul 2 17:09:57 2013
(r252512)
@@ -83,12 +83,17 @@ re_patch(void)
void
open_patch_file(char *filename)
{
+ int nr, nw;
+
if (filename == Nullch || !*filename || strEQ(filename, -)) {
pfp = fopen(TMPPATNAME, w);
if (pfp == Nullfp)
pfatal2(can't create %s, TMPPATNAME);
- while (fgets(buf, buf_size, stdin) != Nullch)
- fputs(buf, pfp);
+ while ((nr = fread(buf, 1, buf_size, stdin)) 0) {
+ nw = fwrite(buf, 1, nr, pfp);
+ if (nr != nw)
+ pfatal2(write error to %s, TMPPATNAME);
+ }
Fclose(pfp);
filename = TMPPATNAME;
}
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: should_yield problem [Was: svn commit: r251322 - head/sys/kern]
On Tue, Jul 2, 2013 at 6:48 PM, Andriy Gapon a...@freebsd.org wrote: on 03/06/2013 20:36 Konstantin Belousov said the following: Author: kib Date: Mon Jun 3 17:36:43 2013 New Revision: 251322 URL: http://svnweb.freebsd.org/changeset/base/251322 Log: Be more generous when donating the current thread time to the owner of the vnode lock while iterating over the free vnode list. Instead of yielding, pause for 1 tick. The change is reported to help in some virtualized environments. Kostik, I've just run into a problem where word generous does not seem to be applicable at all unless I am mistaken about how the code works. While ticks is signed it is used as an unsigned int. td_swvolticks is always derived by ticks, which again will always be used as an unisgned and then the subtractions among the 2 will be consistent. Attilio -- Peace can only be achieved by understanding - A. Einstein ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r252513 - head/gnu/usr.bin/patch
Author: obrien Date: Tue Jul 2 17:17:42 2013 New Revision: 252513 URL: http://svnweb.freebsd.org/changeset/base/252513 Log: Properly handle input lines containing NUL characters such that pgets() accurately fills the read buffer. Callers of pgets() still mis-process the buffer contents if the read line contains NUL characters, but this at least makes pgets() accurate. Modified: head/gnu/usr.bin/patch/pch.c Modified: head/gnu/usr.bin/patch/pch.c == --- head/gnu/usr.bin/patch/pch.cTue Jul 2 17:09:57 2013 (r252512) +++ head/gnu/usr.bin/patch/pch.cTue Jul 2 17:17:42 2013 (r252513) @@ -1181,7 +1181,7 @@ pgets(bool do_indent) indent++; } } - Strncpy(buf, line, len - skipped); + memcpy(buf, line, len - skipped); buf[len - skipped] = '\0'; } return len; ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: should_yield problem [Was: svn commit: r251322 - head/sys/kern]
on 02/07/2013 20:12 Attilio Rao said the following: While ticks is signed it is used as an unsigned int. td_swvolticks is always derived by ticks, which again will always be used as an unisgned and then the subtractions among the 2 will be consistent. So returning to my example where ticks == -1946084020 and td_swvoltick == 0 and (ticks - td-td_swvoltick) == -1946084020. Is this consistent in your opinion? -- Andriy Gapon ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r252514 - head/sbin/swapon
Author: delphij
Date: Tue Jul 2 17:34:34 2013
New Revision: 252514
URL: http://svnweb.freebsd.org/changeset/base/252514
Log:
Plug a memory leak.
Modified:
head/sbin/swapon/swapon.c
Modified: head/sbin/swapon/swapon.c
==
--- head/sbin/swapon/swapon.c Tue Jul 2 17:17:42 2013(r252513)
+++ head/sbin/swapon/swapon.c Tue Jul 2 17:34:34 2013(r252514)
@@ -491,6 +491,7 @@ swap_on_off_md(const char *name, char *m
if (!qflag)
warnx(%s: Device already in use,
vnodefile);
+ free(vnodefile);
return (NULL);
}
error = run_cmd(fd, %s -a -t vnode -n -f %s,
@@ -498,6 +499,7 @@ swap_on_off_md(const char *name, char *m
if (error) {
warnx(mdconfig (attach) error: file=%s,
vnodefile);
+ free(vnodefile);
return (NULL);
}
sfd = fdopen(fd, r);
@@ -537,6 +539,7 @@ swap_on_off_md(const char *name, char *m
if (!qflag)
warnx(md%d on %s: Device already
in use, mdunit, vnodefile);
+ free(vnodefile);
return (NULL);
}
error = run_cmd(NULL, %s -a -t vnode -u %d -f %s,
@@ -544,6 +547,7 @@ swap_on_off_md(const char *name, char *m
if (error) {
warnx(mdconfig (attach) error:
md%d on file=%s, mdunit, vnodefile);
+ free(vnodefile);
return (NULL);
}
}
@@ -557,6 +561,7 @@ swap_on_off_md(const char *name, char *m
if (!qflag)
warnx(md on %s: Device not found,
vnodefile);
+ free(vnodefile);
return (NULL);
}
sfd = fdopen(fd, r);
@@ -599,6 +604,7 @@ swap_on_off_md(const char *name, char *m
if (!qflag)
warnx(md%d on %s: Device not found,
mdunit, vnodefile);
+ free(vnodefile);
return (NULL);
}
}
@@ -622,6 +628,7 @@ err:
fclose(sfd);
if (fd != -1)
close(fd);
+ free(vnodefile);
return (ret);
}
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: should_yield problem [Was: svn commit: r251322 - head/sys/kern]
On Tue, Jul 2, 2013 at 7:24 PM, Andriy Gapon a...@freebsd.org wrote: on 02/07/2013 20:12 Attilio Rao said the following: While ticks is signed it is used as an unsigned int. td_swvolticks is always derived by ticks, which again will always be used as an unisgned and then the subtractions among the 2 will be consistent. So returning to my example where ticks == -1946084020 and td_swvoltick == 0 and (ticks - td-td_swvoltick) == -1946084020. Is this consistent in your opinion? I was just pointing out that the real bug is not in the subtraction itself but on the hogticks comparison. This is because hogticks is not an absolute measurement but it represents really a ticks difference. In my opinion we should define hogticks as an unsigned int and add a small comment saying that it is a differential. This will fix the issue. Thanks, Attilio -- Peace can only be achieved by understanding - A. Einstein ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: should_yield problem [Was: svn commit: r251322 - head/sys/kern]
On 2 July 2013 12:48, Andriy Gapon a...@freebsd.org wrote: I am not sure if the originally reported problem was also caused by should_yield() or if it was something else. But in either case I think that we should fix should_yield. Perhaps (ticks - curthread-td_swvoltick) should be cast to unsigned before comparison? What about just initializing td_swvoltick to ticks at td creation? ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: should_yield problem [Was: svn commit: r251322 - head/sys/kern]
on 02/07/2013 20:52 Attilio Rao said the following: I was just pointing out that the real bug is not in the subtraction itself but on the hogticks comparison. This is because hogticks is not an absolute measurement but it represents really a ticks difference. In my opinion we should define hogticks as an unsigned int and add a small comment saying that it is a differential. This will fix the issue. I think that my original suggestion is equivalently well if not more obvious. This is a common knowledge: http://en.wikipedia.org/wiki/Serial_number_arithmetic#General_Solution distance = (signed)( i1 - i2 ) -- Andriy Gapon ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: should_yield problem [Was: svn commit: r251322 - head/sys/kern]
on 02/07/2013 20:59 Ed Maste said the following: On 2 July 2013 12:48, Andriy Gapon a...@freebsd.org wrote: I am not sure if the originally reported problem was also caused by should_yield() or if it was something else. But in either case I think that we should fix should_yield. Perhaps (ticks - curthread-td_swvoltick) should be cast to unsigned before comparison? What about just initializing td_swvoltick to ticks at td creation? I like this idea. Still I think that distance between two wrapping numbers should be calculated properly even if that mostly would not matter in practice. distance = (signed)( i1 - i2 ) -- Andriy Gapon ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r252515 - head/sys/ufs/ffs
Author: pfg
Date: Tue Jul 2 18:45:37 2013
New Revision: 252515
URL: http://svnweb.freebsd.org/changeset/base/252515
Log:
Style fix: spaces.
Cleanup the incomplete revert.
Reported by: bde
MFC after:4 weeks
Modified:
head/sys/ufs/ffs/ffs_vfsops.c
Modified: head/sys/ufs/ffs/ffs_vfsops.c
==
--- head/sys/ufs/ffs/ffs_vfsops.c Tue Jul 2 17:34:34 2013
(r252514)
+++ head/sys/ufs/ffs/ffs_vfsops.c Tue Jul 2 18:45:37 2013
(r252515)
@@ -1791,7 +1791,7 @@ ffs_vgetf(mp, ino, flags, vpp, ffs_flags
* already have one. This should only happen on old filesystems.
*/
if (ip-i_gen == 0) {
- ip-i_gen = arc4random()/2 + 1;
+ ip-i_gen = arc4random() / 2 + 1;
if ((vp-v_mount-mnt_flag MNT_RDONLY) == 0) {
ip-i_flag |= IN_MODIFIED;
DIP_SET(ip, i_gen, ip-i_gen);
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r252516 - head/usr.sbin/bsnmpd/modules/snmp_hast
Author: trociny
Date: Tue Jul 2 18:54:39 2013
New Revision: 252516
URL: http://svnweb.freebsd.org/changeset/base/252516
Log:
Add a mib for worker process ID.
MFC after:3 days
Modified:
head/usr.sbin/bsnmpd/modules/snmp_hast/BEGEMOT-HAST-MIB.txt
head/usr.sbin/bsnmpd/modules/snmp_hast/hast_snmp.c
head/usr.sbin/bsnmpd/modules/snmp_hast/hast_tree.def
Modified: head/usr.sbin/bsnmpd/modules/snmp_hast/BEGEMOT-HAST-MIB.txt
==
--- head/usr.sbin/bsnmpd/modules/snmp_hast/BEGEMOT-HAST-MIB.txt Tue Jul 2
18:45:37 2013(r252515)
+++ head/usr.sbin/bsnmpd/modules/snmp_hast/BEGEMOT-HAST-MIB.txt Tue Jul 2
18:54:39 2013(r252516)
@@ -57,6 +57,9 @@ begemotHast MODULE-IDENTITY
REVISION 20130413Z
DESCRIPTION
Initial revision.
+REVISION 20130701Z
+DESCRIPTION
+ Added hastResourceWorkerPid.
::= { begemot 220 }
begemotHastObjects OBJECT IDENTIFIER ::= { begemotHast 1 }
@@ -116,7 +119,8 @@ HastResourceEntry ::= SEQUENCE {
hastResourceReadErrors Counter64,
hastResourceWriteErrorsCounter64,
hastResourceDeleteErrors Counter64,
-hastResourceFlushErrorsCounter64
+hastResourceFlushErrorsCounter64,
+hastResourceWorkerPid INTEGER
}
hastResourceIndex OBJECT-TYPE
@@ -295,4 +299,12 @@ hastResourceFlushErrors OBJECT-TYPE
Count of resource local flush operations that failed.
::= { hastResourceEntry 21 }
+hastResourceWorkerPid OBJECT-TYPE
+SYNTAX INTEGER
+MAX-ACCESS read-write
+STATUS current
+DESCRIPTION
+ Worker process ID.
+::= { hastResourceEntry 22 }
+
END
Modified: head/usr.sbin/bsnmpd/modules/snmp_hast/hast_snmp.c
==
--- head/usr.sbin/bsnmpd/modules/snmp_hast/hast_snmp.c Tue Jul 2 18:45:37
2013(r252515)
+++ head/usr.sbin/bsnmpd/modules/snmp_hast/hast_snmp.c Tue Jul 2 18:54:39
2013(r252516)
@@ -78,6 +78,7 @@ struct hast_snmp_resource {
uint64_twrite_errors;
uint64_tdelete_errors;
uint64_tflush_errors;
+ pid_t workerpid;
};
static TAILQ_HEAD(, hast_snmp_resource) resources =
@@ -343,6 +344,7 @@ update_resources(void)
nv_get_uint64(nvout, stat_delete_error%u, i);
res-flush_errors =
nv_get_uint64(nvout, stat_flush_error%u, i);
+ res-workerpid = nv_get_int32(nvout, workerpid%u, i);
TAILQ_INSERT_TAIL(resources, res, link);
}
nv_free(nvout);
@@ -498,6 +500,9 @@ op_hastResourceTable(struct snmp_context
case LEAF_hastResourceFlushErrors:
value-v.counter64 = res-flush_errors;
break;
+ case LEAF_hastResourceWorkerPid:
+ value-v.integer = res-workerpid;
+ break;
default:
ret = SNMP_ERR_RES_UNAVAIL;
break;
Modified: head/usr.sbin/bsnmpd/modules/snmp_hast/hast_tree.def
==
--- head/usr.sbin/bsnmpd/modules/snmp_hast/hast_tree.defTue Jul 2
18:45:37 2013(r252515)
+++ head/usr.sbin/bsnmpd/modules/snmp_hast/hast_tree.defTue Jul 2
18:54:39 2013(r252516)
@@ -59,6 +59,7 @@
(19 hastResourceWriteErrors COUNTER64 GET)
(20 hastResourceDeleteErrors COUNTER64 GET)
(21 hastResourceFlushErrors COUNTER64 GET)
+ (22 hastResourceWorkerPid INTEGER GET)
)
)
)
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r252517 - stable/9/sbin/hastd
Author: trociny
Date: Tue Jul 2 19:27:07 2013
New Revision: 252517
URL: http://svnweb.freebsd.org/changeset/base/252517
Log:
MFC r248294, r248296:
r248294 (pjd):
Delete requests can be larger than MAXPHYS.
r248296 (pjd):
Minor corrections.
Modified:
stable/9/sbin/hastd/hastd.8
stable/9/sbin/hastd/secondary.c
Directory Properties:
stable/9/sbin/hastd/ (props changed)
Modified: stable/9/sbin/hastd/hastd.8
==
--- stable/9/sbin/hastd/hastd.8 Tue Jul 2 18:54:39 2013(r252516)
+++ stable/9/sbin/hastd/hastd.8 Tue Jul 2 19:27:07 2013(r252517)
@@ -51,7 +51,7 @@ Only one machine (cluster node) can acti
This machine is called primary.
The
.Nm
-daemon operates on block level, which makes it transparent for file
+daemon operates on block level, which makes it transparent to file
systems and applications.
.Pp
There is one main
@@ -68,7 +68,7 @@ The exact format is:
hastd: resource name (role)
.Ed
.Pp
-When (and only when)
+If (and only if)
.Nm
operates in primary role for the given resource, corresponding
.Pa /dev/hast/name
@@ -77,8 +77,8 @@ File systems and applications can use th
requests to.
Every write, delete and flush operation
.Dv ( BIO_WRITE , BIO_DELETE , BIO_FLUSH )
-is send to local component and synchronously replicated
-to the remote (secondary) node if it is available.
+is send to local component and replicated to the remote (secondary) node if it
+is available.
Read operations
.Dv ( BIO_READ )
are handled locally unless I/O error occurs or local version of the data
Modified: stable/9/sbin/hastd/secondary.c
==
--- stable/9/sbin/hastd/secondary.c Tue Jul 2 18:54:39 2013
(r252516)
+++ stable/9/sbin/hastd/secondary.c Tue Jul 2 19:27:07 2013
(r252517)
@@ -582,7 +582,7 @@ requnpack(struct hast_resource *res, str
hio-hio_error = EINVAL;
goto end;
}
- if (hio-hio_length MAXPHYS) {
+ if (hio-hio_cmd != HIO_DELETE hio-hio_length MAXPHYS) {
pjdlog_error(Data length is too large (%ju %ju).,
(uintmax_t)hio-hio_length, (uintmax_t)MAXPHYS);
hio-hio_error = EINVAL;
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: svn commit: r252514 - head/sbin/swapon
Le 02/07/2013 19:34, Xin LI a écrit : Author: delphij Date: Tue Jul 2 17:34:34 2013 New Revision: 252514 URL: http://svnweb.freebsd.org/changeset/base/252514 Log: Plug a memory leak. Modified: head/sbin/swapon/swapon.c Modified: head/sbin/swapon/swapon.c == --- head/sbin/swapon/swapon.c Tue Jul 2 17:17:42 2013(r252513) +++ head/sbin/swapon/swapon.c Tue Jul 2 17:34:34 2013(r252514) Hi, Wouldn't be simpler to do a goto somewhere in the code to do the free() and return(NULL) instead of copying those instructions at each case ? It's just a question. -- Florent Peterschmitt | Please: flor...@peterschmitt.fr| * Avoid HTML/RTF in E-mail. +33 (0)6 64 33 97 92 | * Send PDF for documents. http://florent.peterschmitt.fr | Thank you :) signature.asc Description: OpenPGP digital signature
svn commit: r252518 - stable/8/sbin/hastd
Author: trociny
Date: Tue Jul 2 19:28:16 2013
New Revision: 252518
URL: http://svnweb.freebsd.org/changeset/base/252518
Log:
MFC r248294, r248296:
r248294 (pjd):
Delete requests can be larger than MAXPHYS.
r248296 (pjd):
Minor corrections.
Modified:
stable/8/sbin/hastd/hastd.8
stable/8/sbin/hastd/secondary.c
Directory Properties:
stable/8/sbin/hastd/ (props changed)
Modified: stable/8/sbin/hastd/hastd.8
==
--- stable/8/sbin/hastd/hastd.8 Tue Jul 2 19:27:07 2013(r252517)
+++ stable/8/sbin/hastd/hastd.8 Tue Jul 2 19:28:16 2013(r252518)
@@ -51,7 +51,7 @@ Only one machine (cluster node) can acti
This machine is called primary.
The
.Nm
-daemon operates on block level, which makes it transparent for file
+daemon operates on block level, which makes it transparent to file
systems and applications.
.Pp
There is one main
@@ -68,7 +68,7 @@ The exact format is:
hastd: resource name (role)
.Ed
.Pp
-When (and only when)
+If (and only if)
.Nm
operates in primary role for the given resource, corresponding
.Pa /dev/hast/name
@@ -77,8 +77,8 @@ File systems and applications can use th
requests to.
Every write, delete and flush operation
.Dv ( BIO_WRITE , BIO_DELETE , BIO_FLUSH )
-is send to local component and synchronously replicated
-to the remote (secondary) node if it is available.
+is send to local component and replicated to the remote (secondary) node if it
+is available.
Read operations
.Dv ( BIO_READ )
are handled locally unless I/O error occurs or local version of the data
Modified: stable/8/sbin/hastd/secondary.c
==
--- stable/8/sbin/hastd/secondary.c Tue Jul 2 19:27:07 2013
(r252517)
+++ stable/8/sbin/hastd/secondary.c Tue Jul 2 19:28:16 2013
(r252518)
@@ -582,7 +582,7 @@ requnpack(struct hast_resource *res, str
hio-hio_error = EINVAL;
goto end;
}
- if (hio-hio_length MAXPHYS) {
+ if (hio-hio_cmd != HIO_DELETE hio-hio_length MAXPHYS) {
pjdlog_error(Data length is too large (%ju %ju).,
(uintmax_t)hio-hio_length, (uintmax_t)MAXPHYS);
hio-hio_error = EINVAL;
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r252519 - stable/9/sbin/hastctl
Author: trociny Date: Tue Jul 2 19:29:00 2013 New Revision: 252519 URL: http://svnweb.freebsd.org/changeset/base/252519 Log: MFC r248286 (pjd): Removed redundant includes. Modified: stable/9/sbin/hastctl/hastctl.c Directory Properties: stable/9/sbin/hastctl/ (props changed) Modified: stable/9/sbin/hastctl/hastctl.c == --- stable/9/sbin/hastctl/hastctl.c Tue Jul 2 19:28:16 2013 (r252518) +++ stable/9/sbin/hastctl/hastctl.c Tue Jul 2 19:29:00 2013 (r252519) @@ -31,21 +31,11 @@ __FBSDID($FreeBSD$); #include sys/param.h -#include sys/disk.h -#include sys/ioctl.h -#include sys/stat.h -#include sys/sysctl.h #include err.h -#include errno.h -#include fcntl.h #include libutil.h -#include limits.h -#include signal.h #include stdio.h -#include stdlib.h #include string.h -#include sysexits.h #include unistd.h #include activemap.h ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r252520 - stable/8/sbin/hastctl
Author: trociny Date: Tue Jul 2 19:29:45 2013 New Revision: 252520 URL: http://svnweb.freebsd.org/changeset/base/252520 Log: MFC r248286 (pjd): Removed redundant includes. Modified: stable/8/sbin/hastctl/hastctl.c Directory Properties: stable/8/sbin/hastctl/ (props changed) Modified: stable/8/sbin/hastctl/hastctl.c == --- stable/8/sbin/hastctl/hastctl.c Tue Jul 2 19:29:00 2013 (r252519) +++ stable/8/sbin/hastctl/hastctl.c Tue Jul 2 19:29:45 2013 (r252520) @@ -31,21 +31,11 @@ __FBSDID($FreeBSD$); #include sys/param.h -#include sys/disk.h -#include sys/ioctl.h -#include sys/stat.h -#include sys/sysctl.h #include err.h -#include errno.h -#include fcntl.h #include libutil.h -#include limits.h -#include signal.h #include stdio.h -#include stdlib.h #include string.h -#include sysexits.h #include unistd.h #include activemap.h ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r252521 - svnadmin/conf
Author: gavin Date: Tue Jul 2 19:33:19 2013 New Revision: 252521 URL: http://svnweb.freebsd.org/changeset/base/252521 Log: Take into safekeeping the following bits: anchie, ariff, daichi, darrenr, dd, eivind, flz, green, hmp, iedowse, jls, kmacy, kris, le, luoqi, markus, mbr, mlaier, motoyuki, mpp, mr, nork, pb, peadar, rik, rodrigc, rse, sam, sepotvin, simokawa, sson, ssouhlal, suz, syrinx, tanimura, ticso, weongyo, wes, wpaul. All have been idle in the src tree for some time, and have not responded to emails asking whether they are likely to become active again and so wish to retain their commit bit. Remove them, and entries from conf/mentors where necessary. Thank you all for all your work in the past, and feel free to request your commit bit back should you find yourself able to contribute again in the future. Approved by: core Modified: svnadmin/conf/access svnadmin/conf/mentors Modified: svnadmin/conf/access == --- svnadmin/conf/accessTue Jul 2 19:29:45 2013(r252520) +++ svnadmin/conf/accessTue Jul 2 19:33:19 2013(r252521) @@ -25,12 +25,10 @@ akiyama alc alfred ambrisko -anchie andre andreast andrew antoine -ariff asomers attilio avatar @@ -54,15 +52,12 @@ cperciva cracauer csjp cy -daichi -darrenr das davidch davidcs davide davidxu dchagin -dd deischen delphij des @@ -75,7 +70,6 @@ dwmalone dwmalone=freebsd-committers@mat eadler ed edwin -eivind emaste emax eri @@ -83,7 +77,6 @@ erwin fabient fanf fjoe -flz gabor gad gallatin @@ -96,17 +89,14 @@ gleb glebius gnn gonzo -green grehan grog gshapiro freebsd-cvs-committ...@g.gshapiro.net harti hiren -hmp hrs hselasky ian -iedowse imp ivoras iwasaki @@ -124,7 +114,6 @@ jilles jimharris jkim jlh -jls jmallett jmg joerg freebsd-de...@uriah.heep.sax.de @@ -140,21 +129,15 @@ kensmith kevlo kib kientzle -kmacy -kris kuriyama -le lstewart luigi -luoqi marcel marius markj markm -markus mav maxim -mbr mckay mckusick mdf @@ -162,24 +145,17 @@ melifaro miwi mjacob mjg -mlaier mm monthadar -motoyuki mp -mpp -mr n_hibmadevnull neel netchild -nork np nwhitehorn nyan obrien oleg -pb -peadar peter pe...@wemm.org peterj pfg @@ -194,45 +170,33 @@ raj ray rdivacky remko -rik rmacklem rmh rnoland roberto -rodrigc rpaulo rrs -rse rstone ru rwatson bb+lists.freebsd.cvs-committ...@cyrus.watson.org -sam sanpei sbruno scf schweikh scottl se -sepotvin silby -simokawa simon sjg smh sobomax -sson -ssouhlal stas stefanf -suz -syrinx syuu takawata -tanimura theraven thomas thompsa -ticso tijl trasz trhodes @@ -244,12 +208,9 @@ upsups_...@stups.com uqs vanhu versus -weongyo -wes will wkoszek wollman -wpaul yongari zec zont Modified: svnadmin/conf/mentors == --- svnadmin/conf/mentors Tue Jul 2 19:29:45 2013(r252520) +++ svnadmin/conf/mentors Tue Jul 2 19:33:19 2013(r252521) @@ -11,7 +11,6 @@ # Mentee Mentor Optional comment achim scottl Co-mentor: emaste -anchie bz asomersken Co-mentor: gibbs, will benl philip Co-mentor: simon carl jimharris @@ -26,7 +25,6 @@ jonathan rwatson jwdrmacklem miwi rwatson monthadar adrian -nork imp peterj jhb Co-mentor: grog snbdwmalone versus gavin Co-mentor: fjoe ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r252522 - head/sys/arm/conf
Author: andrew Date: Tue Jul 2 19:35:04 2013 New Revision: 252522 URL: http://svnweb.freebsd.org/changeset/base/252522 Log: Enable VFP on Raspberry Pi. This has worked as of r251712. Modified: head/sys/arm/conf/RPI-B Modified: head/sys/arm/conf/RPI-B == --- head/sys/arm/conf/RPI-B Tue Jul 2 19:33:19 2013(r252521) +++ head/sys/arm/conf/RPI-B Tue Jul 2 19:35:04 2013(r252522) @@ -120,3 +120,6 @@ options FDT # handed to kernel via U-Boot and ubldr. #options FDT_DTB_STATIC makeoptions FDT_DTS_FILE=rpi.dts + +device vfp # vfp/neon +optionsARM_VFP_SUPPORT # vfp/neon ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r252523 - head/sbin/mdconfig
Author: delphij Date: Tue Jul 2 19:42:47 2013 New Revision: 252523 URL: http://svnweb.freebsd.org/changeset/base/252523 Log: When listing with -f, skip all memory disks that are not vnode-backed. Noticed by: kevlo MFC after:3 days Modified: head/sbin/mdconfig/mdconfig.c Modified: head/sbin/mdconfig/mdconfig.c == --- head/sbin/mdconfig/mdconfig.c Tue Jul 2 19:35:04 2013 (r252522) +++ head/sbin/mdconfig/mdconfig.c Tue Jul 2 19:42:47 2013 (r252523) @@ -450,7 +450,8 @@ md_list(const char *units, int opt, cons continue; else ffound = 1; - } + } else if (fflag != NULL) + continue; if (nflag strncmp(pp-lg_name, MD_NAME, 2) == 0) printf(%s, pp-lg_name + 2); else ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r252524 - stable/9/sys/cddl/contrib/opensolaris/uts/common/fs/zfs
Author: mm Date: Tue Jul 2 19:44:02 2013 New Revision: 252524 URL: http://svnweb.freebsd.org/changeset/base/252524 Log: MFC r252380: Unbreak zfs jail and zfs unjail (broken in stable/9 since r249643) I missed to register zfs_ioc_jail and zfs_ioc_unjail as legacy ioctl's with the new zfs_ioctl_register_legacy() function. These operations do not modify pools or datasets so there is no need to log them to pool history. Reported by: Alexander Leidinger netch...@freebsd.org on current@ Modified: stable/9/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ioctl.c Directory Properties: stable/9/sys/ (props changed) stable/9/sys/cddl/contrib/opensolaris/ (props changed) Modified: stable/9/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ioctl.c == --- stable/9/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ioctl.c Tue Jul 2 19:42:47 2013(r252523) +++ stable/9/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ioctl.c Tue Jul 2 19:44:02 2013(r252524) @@ -5575,6 +5575,13 @@ zfs_ioctl_init(void) zfs_ioctl_register_dataset_nolog(ZFS_IOC_TMP_SNAPSHOT, zfs_ioc_tmp_snapshot, zfs_secpolicy_tmp_snapshot, POOL_CHECK_SUSPENDED | POOL_CHECK_READONLY); + +#ifdef __FreeBSD__ + zfs_ioctl_register_dataset_nolog(ZFS_IOC_JAIL, zfs_ioc_jail, + zfs_secpolicy_config, POOL_CHECK_NONE); + zfs_ioctl_register_dataset_nolog(ZFS_IOC_UNJAIL, zfs_ioc_unjail, + zfs_secpolicy_config, POOL_CHECK_NONE); +#endif } int ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r252525 - stable/8/sys/cddl/contrib/opensolaris/uts/common/fs/zfs
Author: mm Date: Tue Jul 2 19:45:14 2013 New Revision: 252525 URL: http://svnweb.freebsd.org/changeset/base/252525 Log: MFC r252380: Unbreak zfs jail and zfs unjail (broken in stable/8 since r251757) I missed to register zfs_ioc_jail and zfs_ioc_unjail as legacy ioctl's with the new zfs_ioctl_register_legacy() function. These operations do not modify pools or datasets so there is no need to log them to pool history. Reported by: Alexander Leidinger netch...@freebsd.org on current@ Modified: stable/8/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ioctl.c Directory Properties: stable/8/sys/ (props changed) stable/8/sys/cddl/ (props changed) stable/8/sys/cddl/contrib/opensolaris/ (props changed) Modified: stable/8/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ioctl.c == --- stable/8/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ioctl.c Tue Jul 2 19:44:02 2013(r252524) +++ stable/8/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ioctl.c Tue Jul 2 19:45:14 2013(r252525) @@ -5575,6 +5575,13 @@ zfs_ioctl_init(void) zfs_ioctl_register_dataset_nolog(ZFS_IOC_TMP_SNAPSHOT, zfs_ioc_tmp_snapshot, zfs_secpolicy_tmp_snapshot, POOL_CHECK_SUSPENDED | POOL_CHECK_READONLY); + +#ifdef __FreeBSD__ + zfs_ioctl_register_dataset_nolog(ZFS_IOC_JAIL, zfs_ioc_jail, + zfs_secpolicy_config, POOL_CHECK_NONE); + zfs_ioctl_register_dataset_nolog(ZFS_IOC_UNJAIL, zfs_ioc_unjail, + zfs_secpolicy_config, POOL_CHECK_NONE); +#endif } int ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: RAND_MAX broken
On 02.07.2013 20:33, Bruce Evans wrote: I checked the values returned by rand(). The ACM part works as intended, so it never returns RAND_MAX. It also never returns 0. So the distribution of values in the documented range [0, RAND_MAX] is very non-uniform. It is uniform in [1, RAND_MAX - 1]. To use this algorithm for rand(), 1 should have been subtracted, giving a range of [0, 0x7ffe]. Do you mean [0, 0x7ffd] (assuming 1 subtracted)? See attached patch. I don't see compatibility problems at least from POSIX specs point of view - they don't say something specific about RAND_MAX. -- http://ache.vniz.net/ bitcoin:13fGiNutKNHcVSsgtGQ7bQ5kgUKgEQHn7N --- stdlib.h.bak2013-07-03 00:08:25.0 +0400 +++ stdlib.h2013-07-03 00:10:17.0 +0400 @@ -69,7 +69,7 @@ #defineEXIT_FAILURE1 #defineEXIT_SUCCESS0 -#defineRAND_MAX0x7fff +#defineRAND_MAX0x7ffd __BEGIN_DECLS #ifdef _XLOCALE_H_ --- rand.c.bak 2013-07-03 00:08:00.0 +0400 +++ rand.c 2013-07-03 00:11:33.0 +0400 @@ -75,7 +75,8 @@ x = 16807 * lo - 2836 * hi; if (x 0) x += 0x7fff; - return ((*ctx = x) % ((u_long)RAND_MAX + 1)); + *ctx = x; + return (x - 1); #endif /* !USE_WEAK_SEEDING */ } ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r252526 - head/share/man/man4
Author: eadler Date: Tue Jul 2 20:25:58 2013 New Revision: 252526 URL: http://svnweb.freebsd.org/changeset/base/252526 Log: Remove extranious '.' Submitted by: swild...@dragonflybsd.org Modified: head/share/man/man4/oce.4 Modified: head/share/man/man4/oce.4 == --- head/share/man/man4/oce.4 Tue Jul 2 19:45:14 2013(r252525) +++ head/share/man/man4/oce.4 Tue Jul 2 20:25:58 2013(r252526) @@ -91,7 +91,7 @@ Firmware can be updated by following the .It Copy the below code to a Makefile: .Bd -literal -offset indent -\.KMOD=elxflash +KMOD=elxflash FIRMWS=imagename.ufi:elxflash \.include bsd.kmod.mk .Ed ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r252527 - head/sys/ufs/ffs
Author: mckusick
Date: Tue Jul 2 21:07:08 2013
New Revision: 252527
URL: http://svnweb.freebsd.org/changeset/base/252527
Log:
Make better use of metadata area by avoiding using it for data blocks
that no should no longer immediately follow their indirect blocks.
MFC after:2 weeks
Modified:
head/sys/ufs/ffs/ffs_alloc.c
head/sys/ufs/ffs/ffs_balloc.c
Modified: head/sys/ufs/ffs/ffs_alloc.c
==
--- head/sys/ufs/ffs/ffs_alloc.cTue Jul 2 20:25:58 2013
(r252526)
+++ head/sys/ufs/ffs/ffs_alloc.cTue Jul 2 21:07:08 2013
(r252527)
@@ -1710,7 +1710,7 @@ ffs_alloccgblk(ip, bp, bpref, size)
cgp = (struct cg *)bp-b_data;
blksfree = cg_blksfree(cgp);
if (bpref == 0) {
- bpref = cgp-cg_rotor;
+ bpref = cgbase(fs, cgp-cg_cgx) + cgp-cg_rotor + fs-fs_frag;
} else if ((cgbpref = dtog(fs, bpref)) != cgp-cg_cgx) {
/* map bpref to correct zone in this cg */
if (bpref cgdata(fs, cgbpref))
Modified: head/sys/ufs/ffs/ffs_balloc.c
==
--- head/sys/ufs/ffs/ffs_balloc.c Tue Jul 2 20:25:58 2013
(r252526)
+++ head/sys/ufs/ffs/ffs_balloc.c Tue Jul 2 21:07:08 2013
(r252527)
@@ -299,6 +299,10 @@ retry:
continue;
}
UFS_LOCK(ump);
+ /*
+* If parent indirect has just been allocated, try to cluster
+* immediately following it.
+*/
if (pref == 0)
pref = ffs_blkpref_ufs1(ip, lbn, i - num - 1,
(ufs1_daddr_t *)0);
@@ -368,7 +372,14 @@ retry:
*/
if (nb == 0) {
UFS_LOCK(ump);
- if (pref == 0)
+ /*
+* If allocating metadata at the front of the cylinder
+* group and parent indirect block has just been allocated,
+* then cluster next to it if it is the first indirect in
+* the file. Otherwise it has been allocated in the metadata
+* area, so we want to find our own place out in the data area.
+*/
+ if (pref == 0 || (lbn NDADDR fs-fs_metaspace != 0))
pref = ffs_blkpref_ufs1(ip, lbn, indirs[i].in_off,
bap[0]);
error = ffs_alloc(ip, lbn, pref, (int)fs-fs_bsize,
@@ -850,6 +861,10 @@ retry:
continue;
}
UFS_LOCK(ump);
+ /*
+* If parent indirect has just been allocated, try to cluster
+* immediately following it.
+*/
if (pref == 0)
pref = ffs_blkpref_ufs2(ip, lbn, i - num - 1,
(ufs2_daddr_t *)0);
@@ -920,7 +935,14 @@ retry:
*/
if (nb == 0) {
UFS_LOCK(ump);
- if (pref == 0)
+ /*
+* If allocating metadata at the front of the cylinder
+* group and parent indirect block has just been allocated,
+* then cluster next to it if it is the first indirect in
+* the file. Otherwise it has been allocated in the metadata
+* area, so we want to find our own place out in the data area.
+*/
+ if (pref == 0 || (lbn NDADDR fs-fs_metaspace != 0))
pref = ffs_blkpref_ufs2(ip, lbn, indirs[i].in_off,
bap[0]);
error = ffs_alloc(ip, lbn, pref, (int)fs-fs_bsize,
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: svn commit: r252516 - head/usr.sbin/bsnmpd/modules/snmp_hast
On Tue, Jul 02, 2013 at 06:54:39PM +, Mikolaj Golub wrote:
Author: trociny
Date: Tue Jul 2 18:54:39 2013
New Revision: 252516
URL: http://svnweb.freebsd.org/changeset/base/252516
Log:
Add a mib for worker process ID.
MFC after: 3 days
Modified:
head/usr.sbin/bsnmpd/modules/snmp_hast/BEGEMOT-HAST-MIB.txt
head/usr.sbin/bsnmpd/modules/snmp_hast/hast_snmp.c
head/usr.sbin/bsnmpd/modules/snmp_hast/hast_tree.def
Modified: head/usr.sbin/bsnmpd/modules/snmp_hast/BEGEMOT-HAST-MIB.txt
==
--- head/usr.sbin/bsnmpd/modules/snmp_hast/BEGEMOT-HAST-MIB.txt Tue Jul
2 18:45:37 2013(r252515)
+++ head/usr.sbin/bsnmpd/modules/snmp_hast/BEGEMOT-HAST-MIB.txt Tue Jul
2 18:54:39 2013(r252516)
@@ -57,6 +57,9 @@ begemotHast MODULE-IDENTITY
[snip]
@@ -295,4 +299,12 @@ hastResourceFlushErrors OBJECT-TYPE
Count of resource local flush operations that failed.
::= { hastResourceEntry 21 }
+hastResourceWorkerPid OBJECT-TYPE
+SYNTAX INTEGER
+MAX-ACCESS read-write
I think MAX-ACCESS should be read-only here. The process ID is assigned
when the worker process is created and it makes no sense for a network
management system to change it.
+STATUS current
+DESCRIPTION
+ Worker process ID.
+::= { hastResourceEntry 22 }
+
END
--
Jilles Tjoelker
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r252528 - head/sys/fs/nfsclient
Author: rmacklem
Date: Wed Jul 3 00:19:03 2013
New Revision: 252528
URL: http://svnweb.freebsd.org/changeset/base/252528
Log:
A problem with the old NFS client where large writes to large files
would sometimes result in a corrupted file was reported via email.
This problem appears to have been caused by r251719 (reverting
r251719 fixed the problem). Although I have not been able to
reproduce this problem, I suspect it is caused by another thread
increasing np-n_size after the mtx_unlock(np-n_mtx) but before
the vnode_pager_setsize() call. Since the np-n_mtx mutex serializes
updates to np-n_size, doing the vnode_pager_setsize() with the
mutex locked appears to avoid the problem.
Unfortunately, vnode_pager_setsize() where the new size is smaller,
cannot be called with a mutex held.
This patch returns the semantics to be close to pre-r251719 (actually
pre-r248567, r248581, r248567 for the new client) such that the call to
vnode_pager_setsize() is only delayed until after the mutex is
unlocked when np-n_size is shrinking. Since the file is growing
when being written, I believe this will fix the corruption.
A better solution might be to replace the mutex with a sleep lock,
but that is a non-trivial conversion, so this fix is hoped to be
sufficient in the meantime.
Reported by: David G. Lawrence (d...@dglawrence.com)
Tested by:David G. Lawrence (to be done soon)
Reviewed by: kib
MFC after:1 week
Modified:
head/sys/fs/nfsclient/nfs_clport.c
Modified: head/sys/fs/nfsclient/nfs_clport.c
==
--- head/sys/fs/nfsclient/nfs_clport.c Tue Jul 2 21:07:08 2013
(r252527)
+++ head/sys/fs/nfsclient/nfs_clport.c Wed Jul 3 00:19:03 2013
(r252528)
@@ -433,6 +433,7 @@ nfscl_loadattrcache(struct vnode **vpp,
vap-va_size = np-n_size;
np-n_attrstamp = 0;
KDTRACE_NFS_ATTRCACHE_FLUSH_DONE(vp);
+ vnode_pager_setsize(vp, np-n_size);
} else if (np-n_flag NMODIFIED) {
/*
* We've modified the file: Use the larger
@@ -444,12 +445,22 @@ nfscl_loadattrcache(struct vnode **vpp,
np-n_size = vap-va_size;
np-n_flag |= NSIZECHANGED;
}
+ vnode_pager_setsize(vp, np-n_size);
+ } else if (vap-va_size np-n_size) {
+ /*
+* When shrinking the size, the call to
+* vnode_pager_setsize() cannot be done
+* with the mutex held, so delay it until
+* after the mtx_unlock call.
+*/
+ nsize = np-n_size = vap-va_size;
+ np-n_flag |= NSIZECHANGED;
+ setnsize = 1;
} else {
np-n_size = vap-va_size;
np-n_flag |= NSIZECHANGED;
+ vnode_pager_setsize(vp, np-n_size);
}
- setnsize = 1;
- nsize = vap-va_size;
} else {
np-n_size = vap-va_size;
}
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: svn commit: r252527 - head/sys/ufs/ffs
On Tue, 2 Jul 2013, Kirk McKusick wrote:
Log:
Make better use of metadata area by avoiding using it for data blocks
that no should no longer immediately follow their indirect blocks.
I use the following changes for allocation at the start of cylinder group.
They seem to be related.
% diff -u2 ffs_alloc.c~ ffs_alloc.c
% --- ffs_alloc.c~ Sun Jun 20 12:53:30 2004
% +++ ffs_alloc.c Mon Jan 2 16:52:39 2012
% @@ -1010,5 +1062,6 @@
% * Select the desired position for the next block in a file. The file is
% * logically divided into sections. The first section is composed of the
% - * direct blocks. Each additional section contains fs_maxbpg blocks.
% + * direct blocks and the next fs_maxbpg blocks. Each additional section
% + * contains fs_maxbpg blocks.
% *
% * If no blocks have been allocated in the first section, the policy is to
% @@ -1022,12 +1075,11 @@
% * continues until a cylinder group with greater than the average number
% * of free blocks is found. If the allocation is for the first block in an
% - * indirect block, the information on the previous allocation is unavailable;
% + * indirect block or the previous block is a hole, then the information on
% + * the previous allocation is unavailable;
% * here a best guess is made based upon the logical block number being
% * allocated.
% *
% * If a section is already partially allocated, the policy is to
% - * contiguously allocate fs_maxcontig blocks. The end of one of these
% - * contiguous blocks and the beginning of the next is laid out
% - * contiguously if possible.
% + * allocate blocks contiguously within the section if possible.
% */
% ufs2_daddr_t
IIRC, the comments were out of date before this change. Especially the
final paragraph -- fs_maxcontig is not used for anything.
% @@ -1039,12 +1091,18 @@
% {
% struct fs *fs;
% - int cg;
% - int avgbfree, startcg;
% + int avgbfree, cg, firstsection, newsection, startcg;
BTW, I don't like changing the type of these variables to u_int in
-current. This asks for new sign extension bugs to break the warnings
about old ones. Even 16-bit signed int is almost large enough for cg,
so it is far from necessary to change it from signed to unsigned to
double its range.
%
% fs = ip-i_fs;
% - if (indx % fs-fs_maxbpg == 0 || bap[indx - 1] == 0) {
% - if (lbn NDADDR + NINDIR(fs)) {
% + if (lbn NDADDR + fs-fs_maxbpg) {
% + firstsection = 1;
% + newsection = 0;
% + } else {
% + firstsection = 0;
% + newsection = ((lbn - NDADDR) % fs-fs_maxbpg == 0);
% + }
% + if (indx == 0 || bap[indx - 1] == 0 || newsection) {
% + if (firstsection) {
% cg = ino_to_cg(fs, ip-i_number);
% - return (fs-fs_fpg * cg + fs-fs_frag);
% + return (cgdmin(fs, cg));
% }
% /*
% @@ -1052,8 +1110,17 @@
%* unused data blocks.
%*/
% - if (indx == 0 || bap[indx - 1] == 0)
% - startcg =
% - ino_to_cg(fs, ip-i_number) + lbn / fs-fs_maxbpg;
% - else
% + if (indx == 0 || bap[indx - 1] == 0) {
% + cg = ino_to_cg(fs, ip-i_number) +
% + (lbn - NDADDR) / fs-fs_maxbpg;
% + if (!newsection) {
% + /*
% + * Actually, use our best guess if the
% + * section is not new.
% + */
% + cg %= fs-fs_ncg;
% + return (cgdmin(fs, cg));
% + }
% + startcg = cg;
% + } else
% startcg = dtog(fs, bap[indx - 1]) + 1;
% startcg %= fs-fs_ncg;
% @@ -1062,16 +1129,13 @@
% if (fs-fs_cs(fs, cg).cs_nbfree = avgbfree) {
% fs-fs_cgrotor = cg;
% - return (fs-fs_fpg * cg + fs-fs_frag);
% + return (cgdmin(fs, cg));
% }
% for (cg = 0; cg = startcg; cg++)
% if (fs-fs_cs(fs, cg).cs_nbfree = avgbfree) {
% fs-fs_cgrotor = cg;
% - return (fs-fs_fpg * cg + fs-fs_frag);
% + return (cgdmin(fs, cg));
% }
% return (0);
% }
% - /*
% - * We just always try to lay things out contiguously.
% - */
% return (bap[indx - 1] + fs-fs_frag);
% }
% @@ -1095,5 +1159,5 @@
% if (lbn NDADDR + NINDIR(fs)) {
% cg = ino_to_cg(fs, ip-i_number);
% - return (fs-fs_fpg * cg + fs-fs_frag);
% + return (cgdmin(fs, cg));
% }
% /*
% @@
svn commit: r252532 - stable/9/sys/kern
Author: lstewart
Date: Wed Jul 3 03:36:36 2013
New Revision: 252532
URL: http://svnweb.freebsd.org/changeset/base/252532
Log:
When a previous call to sbsndptr() leaves sb-sb_sndptroff at the start of an
mbuf that was fully consumed by the previous call, the mbuf ptr returned by
the
current call ends up being the previous mbuf in the sb chain to the one that
contains the data we want.
This does not cause any observable issues because the mbuf copy routines
happily
walk the mbuf chain to get to the data at the moff offset, which in this case
means they effectively skip over the mbuf returned by sbsndptr().
We can't adjust sb-sb_sndptr during the previous call for this case because
the
next mbuf in the chain may not exist yet. We therefore need to detect the
condition and make the adjustment during the current call.
Fix by detecting the special case of moff being at the start of the next mbuf
in
the chain and adjust the required accounting variables accordingly.
Reviewed by: andre
Modified:
stable/9/sys/kern/uipc_sockbuf.c
Directory Properties:
stable/9/sys/ (props changed)
Modified: stable/9/sys/kern/uipc_sockbuf.c
==
--- stable/9/sys/kern/uipc_sockbuf.cWed Jul 3 02:14:12 2013
(r252531)
+++ stable/9/sys/kern/uipc_sockbuf.cWed Jul 3 03:36:36 2013
(r252532)
@@ -939,6 +939,13 @@ sbsndptr(struct sockbuf *sb, u_int off,
/* Return closest mbuf in chain for current offset. */
*moff = off - sb-sb_sndptroff;
m = ret = sb-sb_sndptr ? sb-sb_sndptr : sb-sb_mb;
+ if (*moff == m-m_len) {
+ *moff = 0;
+ sb-sb_sndptroff += m-m_len;
+ m = ret = m-m_next;
+ KASSERT(ret-m_len 0,
+ (mbuf %p in sockbuf %p chain has no valid data, ret, sb));
+ }
/* Advance by len to be as close as possible for the next transmit. */
for (off = off - sb-sb_sndptroff + len - 1;
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r252533 - stable/8/sys/kern
Author: lstewart
Date: Wed Jul 3 03:40:06 2013
New Revision: 252533
URL: http://svnweb.freebsd.org/changeset/base/252533
Log:
When a previous call to sbsndptr() leaves sb-sb_sndptroff at the start of an
mbuf that was fully consumed by the previous call, the mbuf ptr returned by
the
current call ends up being the previous mbuf in the sb chain to the one that
contains the data we want.
This does not cause any observable issues because the mbuf copy routines
happily
walk the mbuf chain to get to the data at the moff offset, which in this case
means they effectively skip over the mbuf returned by sbsndptr().
We can't adjust sb-sb_sndptr during the previous call for this case because
the
next mbuf in the chain may not exist yet. We therefore need to detect the
condition and make the adjustment during the current call.
Fix by detecting the special case of moff being at the start of the next mbuf
in
the chain and adjust the required accounting variables accordingly.
Reviewed by: andre
Modified:
stable/8/sys/kern/uipc_sockbuf.c
Directory Properties:
stable/8/sys/ (props changed)
stable/8/sys/kern/ (props changed)
Modified: stable/8/sys/kern/uipc_sockbuf.c
==
--- stable/8/sys/kern/uipc_sockbuf.cWed Jul 3 03:36:36 2013
(r252532)
+++ stable/8/sys/kern/uipc_sockbuf.cWed Jul 3 03:40:06 2013
(r252533)
@@ -939,6 +939,13 @@ sbsndptr(struct sockbuf *sb, u_int off,
/* Return closest mbuf in chain for current offset. */
*moff = off - sb-sb_sndptroff;
m = ret = sb-sb_sndptr ? sb-sb_sndptr : sb-sb_mb;
+ if (*moff == m-m_len) {
+ *moff = 0;
+ sb-sb_sndptroff += m-m_len;
+ m = ret = m-m_next;
+ KASSERT(ret-m_len 0,
+ (mbuf %p in sockbuf %p chain has no valid data, ret, sb));
+ }
/* Advance by len to be as close as possible for the next transmit. */
for (off = off - sb-sb_sndptroff + len - 1;
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: svn commit: r252346 - head/share/man/man9
On 7/3/13 12:00 AM, John Baldwin wrote: On 6/29/13 6:37 AM, Julian Elischer wrote: On 6/29/13 9:16 PM, Julian Elischer wrote: thanks! -If the owner is not currently actually running then the spin step is skipped. +then a thread attempting to acquire the mutex will spin rather than yielding +the processor. Am I wrong in thinking that it will only spin for a short while, eventually yielding? The original text said this but the new text implies it will spin forever. It does not use a spin timeout. I considered adding a separate note to define adaptive spinning generically as it is used by multiple lock types (mutexes, rwlocks, and sx locks). this would be a good idea. ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r252534 - in stable/9/sys: kern sys
Author: lstewart
Date: Wed Jul 3 04:20:36 2013
New Revision: 252534
URL: http://svnweb.freebsd.org/changeset/base/252534
Log:
MFC r251732,251752:
Add support for non-virtualised hhook points, which are uniquely identified by
type and id, as compared to virtualised hook points which are now uniquely
identified by type, id and a vid (which for vimage is the pointer to the vnet
that the hhook resides in).
All hhook_head structs for both virtualised and non-virtualised hook points
coexist in hhook_head_list, and a separate list is maintained for hhook points
within each vnet to simplify some vimage-related housekeeping.
Reviewed by: scottl
Modified:
stable/9/sys/kern/kern_hhook.c
stable/9/sys/sys/hhook.h
Directory Properties:
stable/9/sys/ (props changed)
stable/9/sys/sys/ (props changed)
Modified: stable/9/sys/kern/kern_hhook.c
==
--- stable/9/sys/kern/kern_hhook.c Wed Jul 3 03:40:06 2013
(r252533)
+++ stable/9/sys/kern/kern_hhook.c Wed Jul 3 04:20:36 2013
(r252534)
@@ -61,8 +61,9 @@ struct hhook {
static MALLOC_DEFINE(M_HHOOK, hhook, Helper hooks are linked off hhook_head
lists);
LIST_HEAD(hhookheadhead, hhook_head);
-VNET_DEFINE(struct hhookheadhead, hhook_head_list);
-#defineV_hhook_head_list VNET(hhook_head_list)
+struct hhookheadhead hhook_head_list;
+VNET_DEFINE(struct hhookheadhead, hhook_vhead_list);
+#defineV_hhook_vhead_list VNET(hhook_vhead_list)
static struct mtx hhook_head_list_lock;
MTX_SYSINIT(hhookheadlistlock, hhook_head_list_lock, hhook_head list lock,
@@ -245,13 +246,6 @@ hhook_head_register(int32_t hhook_type,
return (EEXIST);
}
- /* XXXLAS: Need to implement support for non-virtualised hooks. */
- if ((flags HHOOK_HEADISINVNET) == 0) {
- printf(%s: only vnet-style virtualised hooks can be used\n,
- __func__);
- return (EINVAL);
- }
-
tmphhh = malloc(sizeof(struct hhook_head), M_HHOOK,
M_ZERO | ((flags HHOOK_WAITOK) ? M_WAITOK : M_NOWAIT));
@@ -270,14 +264,17 @@ hhook_head_register(int32_t hhook_type,
} else
refcount_init(tmphhh-hhh_refcount, 0);
+ HHHLIST_LOCK();
if (flags HHOOK_HEADISINVNET) {
tmphhh-hhh_flags |= HHH_ISINVNET;
- HHHLIST_LOCK();
- LIST_INSERT_HEAD(V_hhook_head_list, tmphhh, hhh_next);
- HHHLIST_UNLOCK();
- } else {
- /* XXXLAS: Add tmphhh to the non-virtualised list. */
+#ifdef VIMAGE
+ KASSERT(curvnet != NULL, (curvnet is NULL));
+ tmphhh-hhh_vid = (uintptr_t)curvnet;
+ LIST_INSERT_HEAD(V_hhook_vhead_list, tmphhh, hhh_vnext);
+#endif
}
+ LIST_INSERT_HEAD(hhook_head_list, tmphhh, hhh_next);
+ HHHLIST_UNLOCK();
return (0);
}
@@ -290,6 +287,10 @@ hhook_head_destroy(struct hhook_head *hh
HHHLIST_LOCK_ASSERT();
LIST_REMOVE(hhh, hhh_next);
+#ifdef VIMAGE
+ if (hhook_head_is_virtualised(hhh) == HHOOK_HEADISINVNET)
+ LIST_REMOVE(hhh, hhh_vnext);
+#endif
HHH_WLOCK(hhh);
STAILQ_FOREACH_SAFE(tmp, hhh-hhh_hooks, hhk_next, tmp2)
free(tmp, M_HHOOK);
@@ -347,10 +348,17 @@ hhook_head_get(int32_t hhook_type, int32
{
struct hhook_head *hhh;
- /* XXXLAS: Pick hhook_head_list based on hhook_head flags. */
HHHLIST_LOCK();
- LIST_FOREACH(hhh, V_hhook_head_list, hhh_next) {
+ LIST_FOREACH(hhh, hhook_head_list, hhh_next) {
if (hhh-hhh_type == hhook_type hhh-hhh_id == hhook_id) {
+#ifdef VIMAGE
+ if (hhook_head_is_virtualised(hhh) ==
+ HHOOK_HEADISINVNET) {
+ KASSERT(curvnet != NULL, (curvnet is NULL));
+ if (hhh-hhh_vid != (uintptr_t)curvnet)
+ continue;
+ }
+#endif
refcount_acquire(hhh-hhh_refcount);
break;
}
@@ -412,7 +420,7 @@ static void
hhook_vnet_init(const void *unused __unused)
{
- LIST_INIT(V_hhook_head_list);
+ LIST_INIT(V_hhook_vhead_list);
}
/*
@@ -429,7 +437,7 @@ hhook_vnet_uninit(const void *unused __u
* subsystem should have already called hhook_head_deregister().
*/
HHHLIST_LOCK();
- LIST_FOREACH_SAFE(hhh, V_hhook_head_list, hhh_next, tmphhh) {
+ LIST_FOREACH_SAFE(hhh, V_hhook_vhead_list, hhh_vnext, tmphhh) {
printf(%s: hhook_head type=%d, id=%d cleanup required\n,
__func__, hhh-hhh_type, hhh-hhh_id);
hhook_head_destroy(hhh);
@@ -439,7 +447,7 @@ hhook_vnet_uninit(const void *unused __u
/*
- * When a vnet is created and being initialised, init
svn commit: r252535 - in stable/8/sys: kern sys
Author: lstewart
Date: Wed Jul 3 04:21:29 2013
New Revision: 252535
URL: http://svnweb.freebsd.org/changeset/base/252535
Log:
MFC r251732,251752:
Add support for non-virtualised hhook points, which are uniquely identified by
type and id, as compared to virtualised hook points which are now uniquely
identified by type, id and a vid (which for vimage is the pointer to the vnet
that the hhook resides in).
All hhook_head structs for both virtualised and non-virtualised hook points
coexist in hhook_head_list, and a separate list is maintained for hhook points
within each vnet to simplify some vimage-related housekeeping.
Reviewed by: scottl
Modified:
stable/8/sys/kern/kern_hhook.c
stable/8/sys/sys/hhook.h
Directory Properties:
stable/8/sys/ (props changed)
stable/8/sys/kern/ (props changed)
stable/8/sys/sys/ (props changed)
Modified: stable/8/sys/kern/kern_hhook.c
==
--- stable/8/sys/kern/kern_hhook.c Wed Jul 3 04:20:36 2013
(r252534)
+++ stable/8/sys/kern/kern_hhook.c Wed Jul 3 04:21:29 2013
(r252535)
@@ -61,8 +61,9 @@ struct hhook {
static MALLOC_DEFINE(M_HHOOK, hhook, Helper hooks are linked off hhook_head
lists);
LIST_HEAD(hhookheadhead, hhook_head);
-VNET_DEFINE(struct hhookheadhead, hhook_head_list);
-#defineV_hhook_head_list VNET(hhook_head_list)
+struct hhookheadhead hhook_head_list;
+VNET_DEFINE(struct hhookheadhead, hhook_vhead_list);
+#defineV_hhook_vhead_list VNET(hhook_vhead_list)
static struct mtx hhook_head_list_lock;
MTX_SYSINIT(hhookheadlistlock, hhook_head_list_lock, hhook_head list lock,
@@ -245,13 +246,6 @@ hhook_head_register(int32_t hhook_type,
return (EEXIST);
}
- /* XXXLAS: Need to implement support for non-virtualised hooks. */
- if ((flags HHOOK_HEADISINVNET) == 0) {
- printf(%s: only vnet-style virtualised hooks can be used\n,
- __func__);
- return (EINVAL);
- }
-
tmphhh = malloc(sizeof(struct hhook_head), M_HHOOK,
M_ZERO | ((flags HHOOK_WAITOK) ? M_WAITOK : M_NOWAIT));
@@ -270,14 +264,17 @@ hhook_head_register(int32_t hhook_type,
} else
refcount_init(tmphhh-hhh_refcount, 0);
+ HHHLIST_LOCK();
if (flags HHOOK_HEADISINVNET) {
tmphhh-hhh_flags |= HHH_ISINVNET;
- HHHLIST_LOCK();
- LIST_INSERT_HEAD(V_hhook_head_list, tmphhh, hhh_next);
- HHHLIST_UNLOCK();
- } else {
- /* XXXLAS: Add tmphhh to the non-virtualised list. */
+#ifdef VIMAGE
+ KASSERT(curvnet != NULL, (curvnet is NULL));
+ tmphhh-hhh_vid = (uintptr_t)curvnet;
+ LIST_INSERT_HEAD(V_hhook_vhead_list, tmphhh, hhh_vnext);
+#endif
}
+ LIST_INSERT_HEAD(hhook_head_list, tmphhh, hhh_next);
+ HHHLIST_UNLOCK();
return (0);
}
@@ -290,6 +287,10 @@ hhook_head_destroy(struct hhook_head *hh
HHHLIST_LOCK_ASSERT();
LIST_REMOVE(hhh, hhh_next);
+#ifdef VIMAGE
+ if (hhook_head_is_virtualised(hhh) == HHOOK_HEADISINVNET)
+ LIST_REMOVE(hhh, hhh_vnext);
+#endif
HHH_WLOCK(hhh);
STAILQ_FOREACH_SAFE(tmp, hhh-hhh_hooks, hhk_next, tmp2)
free(tmp, M_HHOOK);
@@ -347,10 +348,17 @@ hhook_head_get(int32_t hhook_type, int32
{
struct hhook_head *hhh;
- /* XXXLAS: Pick hhook_head_list based on hhook_head flags. */
HHHLIST_LOCK();
- LIST_FOREACH(hhh, V_hhook_head_list, hhh_next) {
+ LIST_FOREACH(hhh, hhook_head_list, hhh_next) {
if (hhh-hhh_type == hhook_type hhh-hhh_id == hhook_id) {
+#ifdef VIMAGE
+ if (hhook_head_is_virtualised(hhh) ==
+ HHOOK_HEADISINVNET) {
+ KASSERT(curvnet != NULL, (curvnet is NULL));
+ if (hhh-hhh_vid != (uintptr_t)curvnet)
+ continue;
+ }
+#endif
refcount_acquire(hhh-hhh_refcount);
break;
}
@@ -412,7 +420,7 @@ static void
hhook_vnet_init(const void *unused __unused)
{
- LIST_INIT(V_hhook_head_list);
+ LIST_INIT(V_hhook_vhead_list);
}
/*
@@ -429,7 +437,7 @@ hhook_vnet_uninit(const void *unused __u
* subsystem should have already called hhook_head_deregister().
*/
HHHLIST_LOCK();
- LIST_FOREACH_SAFE(hhh, V_hhook_head_list, hhh_next, tmphhh) {
+ LIST_FOREACH_SAFE(hhh, V_hhook_vhead_list, hhh_vnext, tmphhh) {
printf(%s: hhook_head type=%d, id=%d cleanup required\n,
__func__, hhh-hhh_type, hhh-hhh_id);
hhook_head_destroy(hhh);
@@ -439,7 +447,7 @@ hhook_vnet_uninit(const void *unused __u
/*
- * When a vnet
Re: svn commit: r252532 - stable/9/sys/kern
On 07/03/13 13:36, Lawrence Stewart wrote: Author: lstewart Date: Wed Jul 3 03:36:36 2013 New Revision: 252532 URL: http://svnweb.freebsd.org/changeset/base/252532 Log: MFC r251984: When a previous call to sbsndptr() leaves sb-sb_sndptroff at the start of an mbuf that was fully consumed by the previous call, the mbuf ptr returned by the current call ends up being the previous mbuf in the sb chain to the one that contains the data we want. This does not cause any observable issues because the mbuf copy routines happily walk the mbuf chain to get to the data at the moff offset, which in this case means they effectively skip over the mbuf returned by sbsndptr(). We can't adjust sb-sb_sndptr during the previous call for this case because the next mbuf in the chain may not exist yet. We therefore need to detect the condition and make the adjustment during the current call. Fix by detecting the special case of moff being at the start of the next mbuf in the chain and adjust the required accounting variables accordingly. Reviewed by:andre ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: svn commit: r252533 - stable/8/sys/kern
On 07/03/13 13:40, Lawrence Stewart wrote: Author: lstewart Date: Wed Jul 3 03:40:06 2013 New Revision: 252533 URL: http://svnweb.freebsd.org/changeset/base/252533 Log: MFC r251984: When a previous call to sbsndptr() leaves sb-sb_sndptroff at the start of an mbuf that was fully consumed by the previous call, the mbuf ptr returned by the current call ends up being the previous mbuf in the sb chain to the one that contains the data we want. This does not cause any observable issues because the mbuf copy routines happily walk the mbuf chain to get to the data at the moff offset, which in this case means they effectively skip over the mbuf returned by sbsndptr(). We can't adjust sb-sb_sndptr during the previous call for this case because the next mbuf in the chain may not exist yet. We therefore need to detect the condition and make the adjustment during the current call. Fix by detecting the special case of moff being at the start of the next mbuf in the chain and adjust the required accounting variables accordingly. Reviewed by:andre ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: should_yield problem [Was: svn commit: r251322 - head/sys/kern]
On Tue, 2 Jul 2013, Andriy Gapon wrote: on 02/07/2013 20:52 Attilio Rao said the following: I was just pointing out that the real bug is not in the subtraction itself but on the hogticks comparison. This is because hogticks is not an absolute measurement but it represents really a ticks difference. In my opinion we should define hogticks as an unsigned int and add a small comment saying that it is a differential. This will fix the issue. This is sort of backwards. hogticks is a small positive integer. Making it unsigned just asks for sign extension bugs when it is compared with signed values. However, the tick counters are circular counters. Making them signed asks for undefined behaviour when they overflow. I think that my original suggestion is equivalently well if not more obvious. This is a common knowledge: http://en.wikipedia.org/wiki/Serial_number_arithmetic#General_Solution distance = (signed)( i1 - i2 ) With 2's complement and benign overflow, this is no different from 'distance = i1 - i2'. But signed counters are not wanted here. A negative distance, or equivalently with 32-bit unsigned ints, an unsigned difference of = 0x8000 means overflow or an initialization error. Your bug is an initialization error. BTW, -ftrapv is broken in gcc-4.2.1. It seems to have worked in gcc-3.3, but gcc-3.3 generates pessimal code for it, with a call to an extern function even for adding ints. gcc-4.2.1 still generates an declaration of the extern function, but never calls it. The declaration is unnecessary even when it is used. These bugs seem to be fixed in clang, at least on amd64. clang generates an explicit inline check for overflow and executes an unsupported instruction on overflow. The code for the check is good with -O but bad (large and slow) with -O0. These bugs are still large at the hardware level. Overflow checking is handled correctly on x86 only for floating point. In floating point, you can set or clear the mode bit for trapping on overflow, so that programs can be tested without recompiling them and without adding slow runtime checks (the hardware has to pipleline the checks well so that they have low cost). Hardware support for trapping on integer overflow keeps getting worse since programmers never check for overflow. i386 has an into instruction that should have been used instead of jo foo to check for overflow. It should have been easier for hardware to optimize than the branch. But AFAIK, into has never been optimized and has always been slower than the branch, so programmers never used it, so AMD removed it in 64-bit mode. Speed tests for incrementing a volatile local variable on corei7 in 32-bit mode: - -fno-trapv: 6.2 cycles (clang bug: warn about -fno-trapv being unused) - add into: 12.3 cycles (error handling is SIGBUS) - -ftrapv:6.7 cycles (error handling is SIGILL) - mod jo to into: 7.3 cycles (error handling is SIGBUS). Bruce ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: RAND_MAX broken
On Wed, 3 Jul 2013, Andrey Chernov wrote: On 02.07.2013 20:33, Bruce Evans wrote: I checked the values returned by rand(). The ACM part works as intended, so it never returns RAND_MAX. It also never returns 0. So the distribution of values in the documented range [0, RAND_MAX] is very non-uniform. It is uniform in [1, RAND_MAX - 1]. To use this algorithm for rand(), 1 should have been subtracted, giving a range of [0, 0x7ffe]. Do you mean [0, 0x7ffd] (assuming 1 subtracted)? Yes. See attached patch. I don't see compatibility problems at least from POSIX specs point of view - they don't say something specific about RAND_MAX. I like the patch. It is binary compatibility problems and maybe applications assuming that (RAND_MAX + 1) is 0x8000 so as to do magic bit operations with it that may be problems. Bruce ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: svn commit: r252214 - in stable/9: bin/chio sys/cam/scsi sys/sys
On Tue, Jul 02, 2013 at 07:53:33 +0200, Andre Albsmeier wrote: On Tue, 25-Jun-2013 at 23:43:50 +0200, Kenneth D. Merry wrote: Author: ken Date: Tue Jun 25 21:43:49 2013 New Revision: 252214 URL: http://svnweb.freebsd.org/changeset/base/252214 Hi Ken, Log: MFC: 249658, 249701 Update chio(1) and ch(4) to support reporting element designators. This allows mapping a tape drive in a changer (as reported by 'chio status') to a sa(4) driver instance by comparing the serial numbers. The designators can be ASCII (which is printed out directly), binary (which is printed in hex format) or UTF-8, which is printed in either native UTF-8 format if the terminal can support it, or in %XX notation for non-ASCII characters. Thanks to Hiroki Sato hrs@ for the explanation and example UTF-8 printing code. chio.h: Modify the changer_element_status structure to add new fields and definitions from the SMC3r16 spec. Rename the original CHIOGSTATUS ioctl to OCHIOGTATUS and define a new CHIOGSTATUS ioctl. Clean up some tab/space issues. chio.c: For the 'status' subcommand, print the designator field if it is supplied by a device. scsi_ch.h:Add new flags for DVCID and CURDATA to the READ ELEMENT STATUS command structure. Add a read_element_status_device_id structure for the data fields in the new standard. Add new unions, dt_or_obsolete and voltage_devid, to hold and address data from either SCSI-2 or newer devices. scsi_ch.c:Implement support for fetching device IDs with READ ELEMENT STATUS data. Add new arguments to scsi_read_element_status() to allow the user to request the DVCID and CURDATA bits. This broke chio status when talking to my QUALSTAR TLS-8211 library: Jul 2 07:08:22 kern.crit server kernel: (ch0:ahd3:0:1:0): READ ELEMENT STATUS. CDB: b8 10 fd e8 00 01 03 00 04 00 00 00 Jul 2 07:08:22 kern.crit server kernel: (ch0:ahd3:0:1:0): CAM status: SCSI Status Error Jul 2 07:08:22 kern.crit server kernel: (ch0:ahd3:0:1:0): SCSI status: Check Condition Jul 2 07:08:22 kern.crit server kernel: (ch0:ahd3:0:1:0): SCSI sense: ILLEGAL REQUEST asc:24,0 (Invalid field in CDB) Jul 2 07:08:22 kern.crit server kernel: (ch0:ahd3:0:1:0): Command byte 6 is invalid Jul 2 07:08:22 kern.crit server kernel: (ch0:ahd3:0:1:0): Error 22, Unretryable error scsi_cmd-flags (Byte 6) must be zero for this library so I had to use this workaround to make it work again until a better solution is available: --- sys/cam/scsi/scsi_ch.c.ORI2013-06-26 13:38:54.0 +0200 +++ sys/cam/scsi/scsi_ch.c2013-07-02 07:42:24.0 +0200 @@ -1245,8 +1245,8 @@ /* tag_action */ MSG_SIMPLE_Q_TAG, /* voltag */ want_voltags, /* sea */ softc-sc_firsts[chet], - /* dvcid */ 1, - /* curdata */ 1, + /* dvcid */ 0, + /* curdata */ 0, /* count */ 1, /* data_ptr */ data, /* dxfer_len */ 1024, @@ -1284,8 +1284,8 @@ /* voltag */ want_voltags, /* sea */ softc-sc_firsts[chet] + cesr-cesr_element_base, - /* dvcid */ 1, - /* curdata */ 1, + /* dvcid */ 0, + /* curdata */ 0, /* count */ cesr-cesr_element_count, /* data_ptr */ data, /* dxfer_len */ size, -Andre Oops, sorry. We need to check the SCSI version to see whether to set those bits, and also fall back in case it doesn't work. I am on vacation and have very spotty net access. I can't do anything about it until I get back next week. Justin and Alan (CCed) can work on the fix, though. Ken -- Kenneth Merry k...@freebsd.org ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r252536 - in stable/9/sys: kern sys
Author: lstewart
Date: Wed Jul 3 05:28:22 2013
New Revision: 252536
URL: http://svnweb.freebsd.org/changeset/base/252536
Log:
MFC r251770:
Internalise handling of virtualised hook points inside
hhook_{add|remove}_hook_lookup() so that khelp (and other potential API
consumers) do not have to care when they attempt to (un)hook a particular hook
point identified by id and type.
Reviewed by: scottl
Modified:
stable/9/sys/kern/kern_hhook.c
stable/9/sys/kern/kern_khelp.c
stable/9/sys/sys/hhook.h
Directory Properties:
stable/9/sys/ (props changed)
stable/9/sys/sys/ (props changed)
Modified: stable/9/sys/kern/kern_hhook.c
==
--- stable/9/sys/kern/kern_hhook.c Wed Jul 3 04:21:29 2013
(r252535)
+++ stable/9/sys/kern/kern_hhook.c Wed Jul 3 05:28:22 2013
(r252536)
@@ -1,5 +1,5 @@
/*-
- * Copyright (c) 2010 Lawrence Stewart lstew...@freebsd.org
+ * Copyright (c) 2010,2013 Lawrence Stewart lstew...@freebsd.org
* Copyright (c) 2010 The FreeBSD Foundation
* All rights reserved.
*
@@ -69,6 +69,9 @@ static struct mtx hhook_head_list_lock;
MTX_SYSINIT(hhookheadlistlock, hhook_head_list_lock, hhook_head list lock,
MTX_DEF);
+/* Protected by hhook_head_list_lock. */
+static uint32_t n_hhookheads;
+
/* Private function prototypes. */
static void hhook_head_destroy(struct hhook_head *hhh);
@@ -165,21 +168,71 @@ hhook_add_hook(struct hhook_head *hhh, s
}
/*
- * Lookup a helper hook point and register a new helper hook function with it.
+ * Register a helper hook function with a helper hook point (including all
+ * virtual instances of the hook point if it is virtualised).
+ *
+ * The logic is unfortunately far more complex than for
+ * hhook_remove_hook_lookup() because hhook_add_hook() can call malloc() with
+ * M_WAITOK and thus we cannot call hhook_add_hook() with the
+ * hhook_head_list_lock held.
+ *
+ * The logic assembles an array of hhook_head structs that correspond to the
+ * helper hook point being hooked and bumps the refcount on each (all done with
+ * the hhook_head_list_lock held). The hhook_head_list_lock is then dropped,
and
+ * hhook_add_hook() is called and the refcount dropped for each hhook_head
+ * struct in the array.
*/
int
hhook_add_hook_lookup(struct hookinfo *hki, uint32_t flags)
{
- struct hhook_head *hhh;
- int error;
+ struct hhook_head **heads_to_hook, *hhh;
+ int error, i, n_heads_to_hook;
- hhh = hhook_head_get(hki-hook_type, hki-hook_id);
+tryagain:
+ error = i = 0;
+ /*
+* Accessing n_hhookheads without hhook_head_list_lock held opens up a
+* race with hhook_head_register() which we are unlikely to lose, but
+* nonetheless have to cope with - hence the complex goto logic.
+*/
+ n_heads_to_hook = n_hhookheads;
+ heads_to_hook = malloc(n_heads_to_hook * sizeof(struct hhook_head *),
+ M_HHOOK, flags HHOOK_WAITOK ? M_WAITOK : M_NOWAIT);
+ if (heads_to_hook == NULL)
+ return (ENOMEM);
- if (hhh == NULL)
- return (ENOENT);
+ HHHLIST_LOCK();
+ LIST_FOREACH(hhh, hhook_head_list, hhh_next) {
+ if (hhh-hhh_type == hki-hook_type
+ hhh-hhh_id == hki-hook_id) {
+ if (i n_heads_to_hook) {
+ heads_to_hook[i] = hhh;
+
refcount_acquire(heads_to_hook[i]-hhh_refcount);
+ i++;
+ } else {
+ /*
+* We raced with hhook_head_register() which
+* inserted a hhook_head that we need to hook
+* but did not malloc space for. Abort this run
+* and try again.
+*/
+ for (i--; i = 0; i--)
+
refcount_release(heads_to_hook[i]-hhh_refcount);
+ free(heads_to_hook, M_HHOOK);
+ HHHLIST_UNLOCK();
+ goto tryagain;
+ }
+ }
+ }
+ HHHLIST_UNLOCK();
- error = hhook_add_hook(hhh, hki, flags);
- hhook_head_release(hhh);
+ for (i--; i = 0; i--) {
+ if (!error)
+ error = hhook_add_hook(heads_to_hook[i], hki, flags);
+ refcount_release(heads_to_hook[i]-hhh_refcount);
+ }
+
+ free(heads_to_hook, M_HHOOK);
return (error);
}
@@ -211,20 +264,21 @@ hhook_remove_hook(struct hhook_head *hhh
}
/*
- * Lookup a helper hook point and remove a helper hook function from it.
+ * Remove a helper hook function from a helper hook point (including all
+ * virtual instances of the hook point if it
svn commit: r252537 - in stable/8/sys: kern sys
Author: lstewart
Date: Wed Jul 3 05:29:21 2013
New Revision: 252537
URL: http://svnweb.freebsd.org/changeset/base/252537
Log:
MFC r251770:
Internalise handling of virtualised hook points inside
hhook_{add|remove}_hook_lookup() so that khelp (and other potential API
consumers) do not have to care when they attempt to (un)hook a particular hook
point identified by id and type.
Reviewed by: scottl
Modified:
stable/8/sys/kern/kern_hhook.c
stable/8/sys/kern/kern_khelp.c
stable/8/sys/sys/hhook.h
Directory Properties:
stable/8/sys/ (props changed)
stable/8/sys/kern/ (props changed)
stable/8/sys/sys/ (props changed)
Modified: stable/8/sys/kern/kern_hhook.c
==
--- stable/8/sys/kern/kern_hhook.c Wed Jul 3 05:28:22 2013
(r252536)
+++ stable/8/sys/kern/kern_hhook.c Wed Jul 3 05:29:21 2013
(r252537)
@@ -1,5 +1,5 @@
/*-
- * Copyright (c) 2010 Lawrence Stewart lstew...@freebsd.org
+ * Copyright (c) 2010,2013 Lawrence Stewart lstew...@freebsd.org
* Copyright (c) 2010 The FreeBSD Foundation
* All rights reserved.
*
@@ -69,6 +69,9 @@ static struct mtx hhook_head_list_lock;
MTX_SYSINIT(hhookheadlistlock, hhook_head_list_lock, hhook_head list lock,
MTX_DEF);
+/* Protected by hhook_head_list_lock. */
+static uint32_t n_hhookheads;
+
/* Private function prototypes. */
static void hhook_head_destroy(struct hhook_head *hhh);
@@ -165,21 +168,71 @@ hhook_add_hook(struct hhook_head *hhh, s
}
/*
- * Lookup a helper hook point and register a new helper hook function with it.
+ * Register a helper hook function with a helper hook point (including all
+ * virtual instances of the hook point if it is virtualised).
+ *
+ * The logic is unfortunately far more complex than for
+ * hhook_remove_hook_lookup() because hhook_add_hook() can call malloc() with
+ * M_WAITOK and thus we cannot call hhook_add_hook() with the
+ * hhook_head_list_lock held.
+ *
+ * The logic assembles an array of hhook_head structs that correspond to the
+ * helper hook point being hooked and bumps the refcount on each (all done with
+ * the hhook_head_list_lock held). The hhook_head_list_lock is then dropped,
and
+ * hhook_add_hook() is called and the refcount dropped for each hhook_head
+ * struct in the array.
*/
int
hhook_add_hook_lookup(struct hookinfo *hki, uint32_t flags)
{
- struct hhook_head *hhh;
- int error;
+ struct hhook_head **heads_to_hook, *hhh;
+ int error, i, n_heads_to_hook;
- hhh = hhook_head_get(hki-hook_type, hki-hook_id);
+tryagain:
+ error = i = 0;
+ /*
+* Accessing n_hhookheads without hhook_head_list_lock held opens up a
+* race with hhook_head_register() which we are unlikely to lose, but
+* nonetheless have to cope with - hence the complex goto logic.
+*/
+ n_heads_to_hook = n_hhookheads;
+ heads_to_hook = malloc(n_heads_to_hook * sizeof(struct hhook_head *),
+ M_HHOOK, flags HHOOK_WAITOK ? M_WAITOK : M_NOWAIT);
+ if (heads_to_hook == NULL)
+ return (ENOMEM);
- if (hhh == NULL)
- return (ENOENT);
+ HHHLIST_LOCK();
+ LIST_FOREACH(hhh, hhook_head_list, hhh_next) {
+ if (hhh-hhh_type == hki-hook_type
+ hhh-hhh_id == hki-hook_id) {
+ if (i n_heads_to_hook) {
+ heads_to_hook[i] = hhh;
+
refcount_acquire(heads_to_hook[i]-hhh_refcount);
+ i++;
+ } else {
+ /*
+* We raced with hhook_head_register() which
+* inserted a hhook_head that we need to hook
+* but did not malloc space for. Abort this run
+* and try again.
+*/
+ for (i--; i = 0; i--)
+
refcount_release(heads_to_hook[i]-hhh_refcount);
+ free(heads_to_hook, M_HHOOK);
+ HHHLIST_UNLOCK();
+ goto tryagain;
+ }
+ }
+ }
+ HHHLIST_UNLOCK();
- error = hhook_add_hook(hhh, hki, flags);
- hhook_head_release(hhh);
+ for (i--; i = 0; i--) {
+ if (!error)
+ error = hhook_add_hook(heads_to_hook[i], hki, flags);
+ refcount_release(heads_to_hook[i]-hhh_refcount);
+ }
+
+ free(heads_to_hook, M_HHOOK);
return (error);
}
@@ -211,20 +264,21 @@ hhook_remove_hook(struct hhook_head *hhh
}
/*
- * Lookup a helper hook point and remove a helper hook function from it.
+ * Remove a helper hook function from a helper hook point (including all
+ *
Re: svn commit: r252214 - in stable/9: bin/chio sys/cam/scsi sys/sys
On Wed, 03-Jul-2013 at 07:22:36 +0200, Kenneth D. Merry wrote: On Tue, Jul 02, 2013 at 07:53:33 +0200, Andre Albsmeier wrote: On Tue, 25-Jun-2013 at 23:43:50 +0200, Kenneth D. Merry wrote: Author: ken Date: Tue Jun 25 21:43:49 2013 New Revision: 252214 URL: http://svnweb.freebsd.org/changeset/base/252214 Hi Ken, Log: MFC: 249658, 249701 Update chio(1) and ch(4) to support reporting element designators. This allows mapping a tape drive in a changer (as reported by 'chio status') to a sa(4) driver instance by comparing the serial numbers. The designators can be ASCII (which is printed out directly), binary (which is printed in hex format) or UTF-8, which is printed in either native UTF-8 format if the terminal can support it, or in %XX notation for non-ASCII characters. Thanks to Hiroki Sato hrs@ for the explanation and example UTF-8 printing code. chio.h: Modify the changer_element_status structure to add new fields and definitions from the SMC3r16 spec. Rename the original CHIOGSTATUS ioctl to OCHIOGTATUS and define a new CHIOGSTATUS ioctl. Clean up some tab/space issues. chio.c: For the 'status' subcommand, print the designator field if it is supplied by a device. scsi_ch.h:Add new flags for DVCID and CURDATA to the READ ELEMENT STATUS command structure. Add a read_element_status_device_id structure for the data fields in the new standard. Add new unions, dt_or_obsolete and voltage_devid, to hold and address data from either SCSI-2 or newer devices. scsi_ch.c:Implement support for fetching device IDs with READ ELEMENT STATUS data. Add new arguments to scsi_read_element_status() to allow the user to request the DVCID and CURDATA bits. This broke chio status when talking to my QUALSTAR TLS-8211 library: Jul 2 07:08:22 kern.crit server kernel: (ch0:ahd3:0:1:0): READ ELEMENT STATUS. CDB: b8 10 fd e8 00 01 03 00 04 00 00 00 Jul 2 07:08:22 kern.crit server kernel: (ch0:ahd3:0:1:0): CAM status: SCSI Status Error Jul 2 07:08:22 kern.crit server kernel: (ch0:ahd3:0:1:0): SCSI status: Check Condition Jul 2 07:08:22 kern.crit server kernel: (ch0:ahd3:0:1:0): SCSI sense: ILLEGAL REQUEST asc:24,0 (Invalid field in CDB) Jul 2 07:08:22 kern.crit server kernel: (ch0:ahd3:0:1:0): Command byte 6 is invalid Jul 2 07:08:22 kern.crit server kernel: (ch0:ahd3:0:1:0): Error 22, Unretryable error scsi_cmd-flags (Byte 6) must be zero for this library so I had to use this workaround to make it work again until a better solution is available: --- sys/cam/scsi/scsi_ch.c.ORI 2013-06-26 13:38:54.0 +0200 +++ sys/cam/scsi/scsi_ch.c 2013-07-02 07:42:24.0 +0200 @@ -1245,8 +1245,8 @@ /* tag_action */ MSG_SIMPLE_Q_TAG, /* voltag */ want_voltags, /* sea */ softc-sc_firsts[chet], -/* dvcid */ 1, -/* curdata */ 1, +/* dvcid */ 0, +/* curdata */ 0, /* count */ 1, /* data_ptr */ data, /* dxfer_len */ 1024, @@ -1284,8 +1284,8 @@ /* voltag */ want_voltags, /* sea */ softc-sc_firsts[chet] + cesr-cesr_element_base, -/* dvcid */ 1, -/* curdata */ 1, +/* dvcid */ 0, +/* curdata */ 0, /* count */ cesr-cesr_element_count, /* data_ptr */ data, /* dxfer_len */ size, -Andre Oops, sorry. We need to check the SCSI version to see whether to set those bits, and also fall back in case it doesn't work. I am on vacation and have very spotty net access. I can't do anything about it until I get back next week. Justin and Alan (CCed) can work on the fix, though. Take your time, for me it's working right now with the above patch... -Andre Ken -- Kenneth Merry k...@freebsd.org -- Linux: Sozialismus, der nicht funktioniert ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r252538 - stable/8/sys/kern
Author: lstewart
Date: Wed Jul 3 05:47:28 2013
New Revision: 252538
URL: http://svnweb.freebsd.org/changeset/base/252538
Log:
MFC r251774:
Add a private KPI between hhook and khelp that allows khelp modules to insert
hook functions into hhook points which register after the modules were loaded
-
potentially useful during boot or if hhook points are dynamically registered.
Modified:
stable/8/sys/kern/kern_hhook.c
stable/8/sys/kern/kern_khelp.c
Directory Properties:
stable/8/sys/ (props changed)
stable/8/sys/kern/ (props changed)
Modified: stable/8/sys/kern/kern_hhook.c
==
--- stable/8/sys/kern/kern_hhook.c Wed Jul 3 05:29:21 2013
(r252537)
+++ stable/8/sys/kern/kern_hhook.c Wed Jul 3 05:47:28 2013
(r252538)
@@ -74,6 +74,7 @@ static uint32_t n_hhookheads;
/* Private function prototypes. */
static void hhook_head_destroy(struct hhook_head *hhh);
+void khelp_new_hhook_registered(struct hhook_head *hhh, uint32_t flags);
#defineHHHLIST_LOCK() mtx_lock(hhook_head_list_lock)
#defineHHHLIST_UNLOCK() mtx_unlock(hhook_head_list_lock)
@@ -311,12 +312,7 @@ hhook_head_register(int32_t hhook_type,
tmphhh-hhh_nhooks = 0;
STAILQ_INIT(tmphhh-hhh_hooks);
HHH_LOCK_INIT(tmphhh);
-
- if (hhh != NULL) {
- refcount_init(tmphhh-hhh_refcount, 1);
- *hhh = tmphhh;
- } else
- refcount_init(tmphhh-hhh_refcount, 0);
+ refcount_init(tmphhh-hhh_refcount, 1);
HHHLIST_LOCK();
if (flags HHOOK_HEADISINVNET) {
@@ -331,6 +327,13 @@ hhook_head_register(int32_t hhook_type,
n_hhookheads++;
HHHLIST_UNLOCK();
+ khelp_new_hhook_registered(tmphhh, flags);
+
+ if (hhh != NULL)
+ *hhh = tmphhh;
+ else
+ refcount_release(tmphhh-hhh_refcount);
+
return (0);
}
Modified: stable/8/sys/kern/kern_khelp.c
==
--- stable/8/sys/kern/kern_khelp.c Wed Jul 3 05:29:21 2013
(r252537)
+++ stable/8/sys/kern/kern_khelp.c Wed Jul 3 05:47:28 2013
(r252538)
@@ -40,7 +40,6 @@ __FBSDID($FreeBSD$);
#include sys/param.h
#include sys/kernel.h
#include sys/hhook.h
-#include sys/jail.h
#include sys/khelp.h
#include sys/lock.h
#include sys/malloc.h
@@ -52,8 +51,6 @@ __FBSDID($FreeBSD$);
#include sys/rwlock.h
#include sys/systm.h
-#include net/vnet.h
-
static struct rwlock khelp_list_lock;
RW_SYSINIT(khelplistlock, khelp_list_lock, helper list lock);
@@ -61,6 +58,7 @@ static TAILQ_HEAD(helper_head, helper) h
/* Private function prototypes. */
static inline void khelp_remove_osd(struct helper *h, struct osd *hosd);
+void khelp_new_hhook_registered(struct hhook_head *hhh, uint32_t flags);
#defineKHELP_LIST_WLOCK() rw_wlock(khelp_list_lock)
#defineKHELP_LIST_WUNLOCK() rw_wunlock(khelp_list_lock)
@@ -289,6 +287,36 @@ khelp_remove_hhook(struct hookinfo *hki)
return (error);
}
+/*
+ * Private KPI between hhook and khelp that allows khelp modules to insert hook
+ * functions into hhook points which register after the modules were loaded.
+ */
+void
+khelp_new_hhook_registered(struct hhook_head *hhh, uint32_t flags)
+{
+ struct helper *h;
+ int error, i;
+
+ KHELP_LIST_RLOCK();
+ TAILQ_FOREACH(h, helpers, h_next) {
+ for (i = 0; i h-h_nhooks; i++) {
+ if (hhh-hhh_type != h-h_hooks[i].hook_type ||
+ hhh-hhh_id != h-h_hooks[i].hook_id)
+ continue;
+ error = hhook_add_hook(hhh, h-h_hooks[i], flags);
+ if (error) {
+ printf(%s: \%s\ khelp module unable to
+ hook type %d id %d due to error %d\n,
+ __func__, h-h_name,
+ h-h_hooks[i].hook_type,
+ h-h_hooks[i].hook_id, error);
+ error = 0;
+ }
+ }
+ }
+ KHELP_LIST_RUNLOCK();
+}
+
int
khelp_modevent(module_t mod, int event_type, void *data)
{
@@ -348,95 +376,3 @@ khelp_modevent(module_t mod, int event_t
return (error);
}
-
-/*
- * This function is called in two separate situations:
- *
- * - When the kernel is booting, it is called directly by the SYSINIT framework
- * to allow Khelp modules which were compiled into the kernel or loaded by the
- * boot loader to insert their non-virtualised hook functions into the kernel.
- *
- * - When the kernel is booting or a vnet is created, this function is also
- * called indirectly through khelp_vnet_init() by the vnet initialisation code.
- * In this situation, Khelp modules are able to insert their virtualised
svn commit: r252539 - stable/9/sys/kern
Author: lstewart
Date: Wed Jul 3 05:48:46 2013
New Revision: 252539
URL: http://svnweb.freebsd.org/changeset/base/252539
Log:
MFC r251774:
Add a private KPI between hhook and khelp that allows khelp modules to insert
hook functions into hhook points which register after the modules were loaded
-
potentially useful during boot or if hhook points are dynamically registered.
Modified:
stable/9/sys/kern/kern_hhook.c
stable/9/sys/kern/kern_khelp.c
Directory Properties:
stable/9/sys/ (props changed)
Modified: stable/9/sys/kern/kern_hhook.c
==
--- stable/9/sys/kern/kern_hhook.c Wed Jul 3 05:47:28 2013
(r252538)
+++ stable/9/sys/kern/kern_hhook.c Wed Jul 3 05:48:46 2013
(r252539)
@@ -74,6 +74,7 @@ static uint32_t n_hhookheads;
/* Private function prototypes. */
static void hhook_head_destroy(struct hhook_head *hhh);
+void khelp_new_hhook_registered(struct hhook_head *hhh, uint32_t flags);
#defineHHHLIST_LOCK() mtx_lock(hhook_head_list_lock)
#defineHHHLIST_UNLOCK() mtx_unlock(hhook_head_list_lock)
@@ -311,12 +312,7 @@ hhook_head_register(int32_t hhook_type,
tmphhh-hhh_nhooks = 0;
STAILQ_INIT(tmphhh-hhh_hooks);
HHH_LOCK_INIT(tmphhh);
-
- if (hhh != NULL) {
- refcount_init(tmphhh-hhh_refcount, 1);
- *hhh = tmphhh;
- } else
- refcount_init(tmphhh-hhh_refcount, 0);
+ refcount_init(tmphhh-hhh_refcount, 1);
HHHLIST_LOCK();
if (flags HHOOK_HEADISINVNET) {
@@ -331,6 +327,13 @@ hhook_head_register(int32_t hhook_type,
n_hhookheads++;
HHHLIST_UNLOCK();
+ khelp_new_hhook_registered(tmphhh, flags);
+
+ if (hhh != NULL)
+ *hhh = tmphhh;
+ else
+ refcount_release(tmphhh-hhh_refcount);
+
return (0);
}
Modified: stable/9/sys/kern/kern_khelp.c
==
--- stable/9/sys/kern/kern_khelp.c Wed Jul 3 05:47:28 2013
(r252538)
+++ stable/9/sys/kern/kern_khelp.c Wed Jul 3 05:48:46 2013
(r252539)
@@ -40,7 +40,6 @@ __FBSDID($FreeBSD$);
#include sys/param.h
#include sys/kernel.h
#include sys/hhook.h
-#include sys/jail.h
#include sys/khelp.h
#include sys/lock.h
#include sys/malloc.h
@@ -52,8 +51,6 @@ __FBSDID($FreeBSD$);
#include sys/rwlock.h
#include sys/systm.h
-#include net/vnet.h
-
static struct rwlock khelp_list_lock;
RW_SYSINIT(khelplistlock, khelp_list_lock, helper list lock);
@@ -61,6 +58,7 @@ static TAILQ_HEAD(helper_head, helper) h
/* Private function prototypes. */
static inline void khelp_remove_osd(struct helper *h, struct osd *hosd);
+void khelp_new_hhook_registered(struct hhook_head *hhh, uint32_t flags);
#defineKHELP_LIST_WLOCK() rw_wlock(khelp_list_lock)
#defineKHELP_LIST_WUNLOCK() rw_wunlock(khelp_list_lock)
@@ -289,6 +287,36 @@ khelp_remove_hhook(struct hookinfo *hki)
return (error);
}
+/*
+ * Private KPI between hhook and khelp that allows khelp modules to insert hook
+ * functions into hhook points which register after the modules were loaded.
+ */
+void
+khelp_new_hhook_registered(struct hhook_head *hhh, uint32_t flags)
+{
+ struct helper *h;
+ int error, i;
+
+ KHELP_LIST_RLOCK();
+ TAILQ_FOREACH(h, helpers, h_next) {
+ for (i = 0; i h-h_nhooks; i++) {
+ if (hhh-hhh_type != h-h_hooks[i].hook_type ||
+ hhh-hhh_id != h-h_hooks[i].hook_id)
+ continue;
+ error = hhook_add_hook(hhh, h-h_hooks[i], flags);
+ if (error) {
+ printf(%s: \%s\ khelp module unable to
+ hook type %d id %d due to error %d\n,
+ __func__, h-h_name,
+ h-h_hooks[i].hook_type,
+ h-h_hooks[i].hook_id, error);
+ error = 0;
+ }
+ }
+ }
+ KHELP_LIST_RUNLOCK();
+}
+
int
khelp_modevent(module_t mod, int event_type, void *data)
{
@@ -348,95 +376,3 @@ khelp_modevent(module_t mod, int event_t
return (error);
}
-
-/*
- * This function is called in two separate situations:
- *
- * - When the kernel is booting, it is called directly by the SYSINIT framework
- * to allow Khelp modules which were compiled into the kernel or loaded by the
- * boot loader to insert their non-virtualised hook functions into the kernel.
- *
- * - When the kernel is booting or a vnet is created, this function is also
- * called indirectly through khelp_vnet_init() by the vnet initialisation code.
- * In this situation, Khelp modules are able to insert their virtualised hook
- * functions into the virtualised
svn commit: r252540 - head/usr.sbin/bsnmpd/modules/snmp_hast
Author: trociny Date: Wed Jul 3 05:54:23 2013 New Revision: 252540 URL: http://svnweb.freebsd.org/changeset/base/252540 Log: hastResourceWorkerPid is read-only. Submitted by: jilles MFC after:3 days Modified: head/usr.sbin/bsnmpd/modules/snmp_hast/BEGEMOT-HAST-MIB.txt Modified: head/usr.sbin/bsnmpd/modules/snmp_hast/BEGEMOT-HAST-MIB.txt == --- head/usr.sbin/bsnmpd/modules/snmp_hast/BEGEMOT-HAST-MIB.txt Wed Jul 3 05:48:46 2013(r252539) +++ head/usr.sbin/bsnmpd/modules/snmp_hast/BEGEMOT-HAST-MIB.txt Wed Jul 3 05:54:23 2013(r252540) @@ -301,7 +301,7 @@ hastResourceFlushErrors OBJECT-TYPE hastResourceWorkerPid OBJECT-TYPE SYNTAX INTEGER -MAX-ACCESS read-write +MAX-ACCESS read-only STATUS current DESCRIPTION Worker process ID. ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: svn commit: r252516 - head/usr.sbin/bsnmpd/modules/snmp_hast
On Tue, Jul 02, 2013 at 11:40:50PM +0200, Jilles Tjoelker wrote: +hastResourceWorkerPid OBJECT-TYPE +SYNTAX INTEGER +MAX-ACCESS read-write I think MAX-ACCESS should be read-only here. The process ID is assigned when the worker process is created and it makes no sense for a network management system to change it. Yes, it was intended to be read-only. Fixed. Thank you for spotting this. -- Mikolaj Golub ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
