Re: svn commit: r323329 - head/sys/sys
On Fri, 8 Sep 2017, Ngie Cooper (yaneurabeya) wrote: On Sep 8, 2017, at 13:09, Mateusz Guzik wrote: ... Log: Allow __builtin_memset instead of bzero for small buffers of known size This change breaks all platforms that use gcc. If it???s not reverted in the next 3 hours, I???ll revert it. Not all. It works for me. My version of it is 15-20 years old and never always worked on old versions of FreeBSD The problem seems to be with inlining parameters. These are fairly broken by default in gcc-4.2.1, and the overrides in kern/pre.mk only work in some cases. -Os is completely broken (with the overrides, it at best gives a 20% pessimization for space, and with other overrides inlining tends to fail). I recently found a workaround: compile with -fno-inline-functions. So my full compiler flags are CONF_CFLAGS=[-march=i386] -Os -fno-inline-functions -fno-inline-functions-called-once. This should reduce problems with inlining. However, kernels still build with the bad default flags of -O2. Please fix ASAP if possible: https://ci.freebsd.org/job/FreeBSD-head-riscv64-build/3668/console . This also shows problems with uninitialized variables. But -Wno-uninitialized is the default for gcc-4.2.1. This breaks lots of warnings Maybe you have different nonstandard options to avoid the breakage. -Wno-uninitialized is set in kern.mk, and also in kern.pre.mk for cddl. (CDDL_CFLAGS apparerently gives a dumbed down subset of CWARNFLAGS to break many more warnings.) There is a problem getting flags passed to all Makefiles. CONF_CFLAGS (and all other makeoptions in config files) is broken for modules (since macros in the kernel Makefile are not inherited). This feature prevents me using modules even if I forget I don't like them. Flags in kern.mk are only broken for boot Makefiles which neglect to include kern.mk (that is, for all current boot Makefiles). Bruce___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323350 - head/sys/ofed/drivers/infiniband/core
Author: hselasky
Date: Sat Sep 9 06:34:20 2017
New Revision: 323350
URL: https://svnweb.freebsd.org/changeset/base/323350
Log:
Remove unsafe access to the LinuxKPI file structure from ibcore.
selwakeup() is now done by the wake_up() family of functions.
MFC after:1 week
Sponsored by: Mellanox Technologies
Modified:
head/sys/ofed/drivers/infiniband/core/ucm.c
head/sys/ofed/drivers/infiniband/core/ucma.c
head/sys/ofed/drivers/infiniband/core/user_mad.c
head/sys/ofed/drivers/infiniband/core/uverbs_main.c
Modified: head/sys/ofed/drivers/infiniband/core/ucm.c
==
--- head/sys/ofed/drivers/infiniband/core/ucm.c Sat Sep 9 06:29:29 2017
(r323349)
+++ head/sys/ofed/drivers/infiniband/core/ucm.c Sat Sep 9 06:34:20 2017
(r323350)
@@ -378,8 +378,6 @@ static int ib_ucm_event_handler(struct ib_cm_id *cm_id
list_add_tail(&uevent->file_list, &ctx->file->events);
list_add_tail(&uevent->ctx_list, &ctx->events);
wake_up_interruptible(&ctx->file->poll_wait);
- if (ctx->file->filp)
- selwakeup(&ctx->file->filp->f_selinfo);
mutex_unlock(&ctx->file->file_mutex);
return 0;
Modified: head/sys/ofed/drivers/infiniband/core/ucma.c
==
--- head/sys/ofed/drivers/infiniband/core/ucma.cSat Sep 9 06:29:29
2017(r323349)
+++ head/sys/ofed/drivers/infiniband/core/ucma.cSat Sep 9 06:34:20
2017(r323350)
@@ -287,8 +287,6 @@ static int ucma_event_handler(struct rdma_cm_id *cm_id
list_add_tail(&uevent->list, &ctx->file->event_list);
wake_up_interruptible(&ctx->file->poll_wait);
- if (ctx->file->filp)
- selwakeup(&ctx->file->filp->f_selinfo);
out:
mutex_unlock(&ctx->file->mut);
return ret;
Modified: head/sys/ofed/drivers/infiniband/core/user_mad.c
==
--- head/sys/ofed/drivers/infiniband/core/user_mad.cSat Sep 9 06:29:29
2017(r323349)
+++ head/sys/ofed/drivers/infiniband/core/user_mad.cSat Sep 9 06:34:20
2017(r323350)
@@ -236,7 +236,6 @@ static int queue_packet(struct ib_umad_file *file,
packet->mad.hdr.id++)
if (agent == __get_agent(file, packet->mad.hdr.id)) {
list_add_tail(&packet->list, &file->recv_list);
- selwakeup(&file->filp->f_selinfo);
wake_up_interruptible(&file->recv_wait);
ret = 0;
break;
Modified: head/sys/ofed/drivers/infiniband/core/uverbs_main.c
==
--- head/sys/ofed/drivers/infiniband/core/uverbs_main.c Sat Sep 9 06:29:29
2017(r323349)
+++ head/sys/ofed/drivers/infiniband/core/uverbs_main.c Sat Sep 9 06:34:20
2017(r323350)
@@ -519,8 +519,6 @@ void ib_uverbs_comp_handler(struct ib_cq *cq, void *cq
spin_unlock_irqrestore(&file->lock, flags);
wake_up_interruptible(&file->poll_wait);
- if (file->filp)
- selwakeup(&file->filp->f_selinfo);
kill_fasync(&file->async_queue, SIGIO, POLL_IN);
}
@@ -554,8 +552,6 @@ static void ib_uverbs_async_handler(struct ib_uverbs_f
spin_unlock_irqrestore(&file->async_file->lock, flags);
wake_up_interruptible(&file->async_file->poll_wait);
- if (file->async_file->filp)
- selwakeup(&file->async_file->filp->f_selinfo);
kill_fasync(&file->async_file->async_queue, SIGIO, POLL_IN);
}
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323349 - in head/sys: compat/linuxkpi/common/include/linux compat/linuxkpi/common/src sys
Author: hselasky
Date: Sat Sep 9 06:29:29 2017
New Revision: 323349
URL: https://svnweb.freebsd.org/changeset/base/323349
Log:
Properly implement poll_wait() in the LinuxKPI. This prevents direct
use of the linux_poll_wakeup() function from unsafe contexts, which
can lead to use-after-free issues.
Instead of calling linux_poll_wakeup() directly use the wake_up()
family of functions in the LinuxKPI to do this.
Bump the FreeBSD version to force recompilation of external kernel modules.
MFC after:1 week
Sponsored by: Mellanox Technologies
Modified:
head/sys/compat/linuxkpi/common/include/linux/fs.h
head/sys/compat/linuxkpi/common/include/linux/poll.h
head/sys/compat/linuxkpi/common/src/linux_compat.c
head/sys/sys/param.h
Modified: head/sys/compat/linuxkpi/common/include/linux/fs.h
==
--- head/sys/compat/linuxkpi/common/include/linux/fs.h Sat Sep 9 06:24:21
2017(r323348)
+++ head/sys/compat/linuxkpi/common/include/linux/fs.h Sat Sep 9 06:29:29
2017(r323349)
@@ -72,6 +72,17 @@ struct dentry {
struct file_operations;
+struct linux_file_wait_queue {
+ struct wait_queue wq;
+ struct wait_queue_head *wqh;
+ atomic_t state;
+#defineLINUX_FWQ_STATE_INIT 0
+#defineLINUX_FWQ_STATE_NOT_READY 1
+#defineLINUX_FWQ_STATE_QUEUED 2
+#defineLINUX_FWQ_STATE_READY 3
+#defineLINUX_FWQ_STATE_MAX 4
+};
+
struct linux_file {
struct file *_file;
const struct file_operations*f_op;
@@ -97,6 +108,7 @@ struct linux_file {
#defineLINUX_KQ_FLAG_NEED_WRITE (1 << 3)
/* protects f_selinfo.si_note */
spinlock_t f_kqlock;
+ struct linux_file_wait_queue f_wait_queue;
};
#definefilelinux_file
Modified: head/sys/compat/linuxkpi/common/include/linux/poll.h
==
--- head/sys/compat/linuxkpi/common/include/linux/poll.hSat Sep 9
06:24:21 2017(r323348)
+++ head/sys/compat/linuxkpi/common/include/linux/poll.hSat Sep 9
06:29:29 2017(r323349)
@@ -40,11 +40,8 @@
typedef struct poll_table_struct {
} poll_table;
-static inline void
-poll_wait(struct linux_file *filp, wait_queue_head_t *wait_address, poll_table
*p)
-{
- /* NOP */
-}
+extern void linux_poll_wait(struct linux_file *, wait_queue_head_t *,
poll_table *);
+#definepoll_wait(...) linux_poll_wait(__VA_ARGS__)
extern void linux_poll_wakeup(struct linux_file *);
Modified: head/sys/compat/linuxkpi/common/src/linux_compat.c
==
--- head/sys/compat/linuxkpi/common/src/linux_compat.c Sat Sep 9 06:24:21
2017(r323348)
+++ head/sys/compat/linuxkpi/common/src/linux_compat.c Sat Sep 9 06:29:29
2017(r323349)
@@ -1023,10 +1023,9 @@ linux_dev_poll(struct cdev *dev, int events, struct th
file = td->td_fpop;
filp->f_flags = file->f_flag;
linux_set_current(td);
- if (filp->f_op->poll != NULL) {
- selrecord(td, &filp->f_selinfo);
+ if (filp->f_op->poll != NULL)
revents = filp->f_op->poll(filp, NULL) & events;
- } else
+ else
revents = 0;
return (revents);
@@ -1034,7 +1033,93 @@ error:
return (events & (POLLHUP|POLLIN|POLLRDNORM|POLLOUT|POLLWRNORM));
}
+/*
+ * This function atomically updates the poll wakeup state and returns
+ * the previous state at the time of update.
+ */
+static uint8_t
+linux_poll_wakeup_state(atomic_t *v, const uint8_t *pstate)
+{
+ int c, old;
+
+ c = v->counter;
+
+ while ((old = atomic_cmpxchg(v, c, pstate[c])) != c)
+ c = old;
+
+ return (c);
+}
+
+
+static int
+linux_poll_wakeup_callback(wait_queue_t *wq, unsigned int wq_state, int flags,
void *key)
+{
+ static const uint8_t state[LINUX_FWQ_STATE_MAX] = {
+ [LINUX_FWQ_STATE_INIT] = LINUX_FWQ_STATE_INIT, /* NOP */
+ [LINUX_FWQ_STATE_NOT_READY] = LINUX_FWQ_STATE_NOT_READY, /* NOP
*/
+ [LINUX_FWQ_STATE_QUEUED] = LINUX_FWQ_STATE_READY,
+ [LINUX_FWQ_STATE_READY] = LINUX_FWQ_STATE_READY, /* NOP */
+ };
+ struct linux_file *filp = container_of(wq, struct linux_file,
f_wait_queue.wq);
+
+ switch (linux_poll_wakeup_state(&filp->f_wait_queue.state, state)) {
+ case LINUX_FWQ_STATE_QUEUED:
+ linux_poll_wakeup(filp);
+ return (1);
+ default:
+ return (0);
+ }
+}
+
void
+linux_poll_wait(struct linux_file *filp, wait_queue_head_t *wqh, poll_table *p)
+{
+ static const uint8_t state[LINUX_FWQ_STATE_MAX] = {
+ [LINUX_FWQ_STATE_INIT] = LINUX_FWQ_STATE_NOT_READY,
+ [LINUX_FWQ_STATE_NOT_READY] = LINUX_FWQ_ST
svn commit: r323348 - head/usr.bin/tail/tests
Author: ngie
Date: Sat Sep 9 06:24:21 2017
New Revision: 323348
URL: https://svnweb.freebsd.org/changeset/base/323348
Log:
Unbreak :broken_pipe
- Capture exit code in pipeline and test in output.
- Drop awk use in favor of `sleep 2`. This helps guarantee the EPIPE
behavior without the potential race.
Modified:
head/usr.bin/tail/tests/tail_test.sh
Modified: head/usr.bin/tail/tests/tail_test.sh
==
--- head/usr.bin/tail/tests/tail_test.shSat Sep 9 06:04:05 2017
(r323347)
+++ head/usr.bin/tail/tests/tail_test.shSat Sep 9 06:24:21 2017
(r323348)
@@ -222,10 +222,10 @@ broken_pipe_head()
}
broken_pipe_body()
{
- atf_expect_fail "Can't seem to get testcase to work in test
environment. Reproduces easily in interactive shell."
-
atf_check -o save:ints seq -f '%128g' 1 1000
- atf_check -s exit:1 -o ignore -e "inline:tail: stdout" tail -n 856 ints
| awk '{ exit }'
+ atf_check -s ignore \
+ -e "inline:tail: stdout\nexit code: 1\n" \
+ -x '(tail -n 856 ints; echo exit code: $? >&2) | sleep 2'
}
@@ -243,5 +243,5 @@ atf_init_test_cases()
atf_add_test_case longfile_rc135782
atf_add_test_case longfile_rc145782_longlines
atf_add_test_case longfile_rn2500
- #atf_add_test_case broken_pipe
+ atf_add_test_case broken_pipe
}
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323347 - head/sys/compat/linuxkpi/common/include/linux
Author: hselasky
Date: Sat Sep 9 06:04:05 2017
New Revision: 323347
URL: https://svnweb.freebsd.org/changeset/base/323347
Log:
Add more sanity checks to linux_fget() in the LinuxKPI. This prevents
returning pointers to file descriptors which were not created by the
LinuxKPI.
MFC after:1 week
Sponsored by: Mellanox Technologies
Modified:
head/sys/compat/linuxkpi/common/include/linux/file.h
Modified: head/sys/compat/linuxkpi/common/include/linux/file.h
==
--- head/sys/compat/linuxkpi/common/include/linux/file.hSat Sep 9
05:56:04 2017(r323346)
+++ head/sys/compat/linuxkpi/common/include/linux/file.hSat Sep 9
06:04:05 2017(r323347)
@@ -53,11 +53,18 @@ linux_fget(unsigned int fd)
cap_rights_t rights;
struct file *file;
+ /* lookup file pointer by file descriptor index */
if (fget_unlocked(curthread->td_proc->p_fd, fd,
- cap_rights_init(&rights), &file, NULL) != 0) {
+ cap_rights_init(&rights), &file, NULL) != 0)
return (NULL);
+
+ /* check if file handle really belongs to us */
+ if (file->f_data == NULL ||
+ file->f_ops != &linuxfileops) {
+ fdrop(file, curthread);
+ return (NULL);
}
- return (struct linux_file *)file->f_data;
+ return ((struct linux_file *)file->f_data);
}
extern void linux_file_free(struct linux_file *filp);
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323346 - in head/sys: powerpc/powerpc riscv/riscv
Author: mjg
Date: Sat Sep 9 05:56:04 2017
New Revision: 323346
URL: https://svnweb.freebsd.org/changeset/base/323346
Log:
Fix riscv and powerpc compilation after r323329.
On these archs bzero is a C function, which triggers a compilation error
as the compiler tries to expand the macro.
Modified:
head/sys/powerpc/powerpc/machdep.c
head/sys/riscv/riscv/machdep.c
Modified: head/sys/powerpc/powerpc/machdep.c
==
--- head/sys/powerpc/powerpc/machdep.c Sat Sep 9 05:50:47 2017
(r323345)
+++ head/sys/powerpc/powerpc/machdep.c Sat Sep 9 05:56:04 2017
(r323346)
@@ -417,43 +417,6 @@ powerpc_init(vm_offset_t fdt, vm_offset_t toc, vm_offs
(sizeof(struct callframe) - 3*sizeof(register_t))) & ~15UL);
}
-void
-bzero(void *buf, size_t len)
-{
- caddr_t p;
-
- p = buf;
-
- while (((vm_offset_t) p & (sizeof(u_long) - 1)) && len) {
- *p++ = 0;
- len--;
- }
-
- while (len >= sizeof(u_long) * 8) {
- *(u_long*) p = 0;
- *((u_long*) p + 1) = 0;
- *((u_long*) p + 2) = 0;
- *((u_long*) p + 3) = 0;
- len -= sizeof(u_long) * 8;
- *((u_long*) p + 4) = 0;
- *((u_long*) p + 5) = 0;
- *((u_long*) p + 6) = 0;
- *((u_long*) p + 7) = 0;
- p += sizeof(u_long) * 8;
- }
-
- while (len >= sizeof(u_long)) {
- *(u_long*) p = 0;
- len -= sizeof(u_long);
- p += sizeof(u_long);
- }
-
- while (len) {
- *p++ = 0;
- len--;
- }
-}
-
/*
* Flush the D-cache for non-DMA I/O so that the I-cache can
* be made coherent later.
@@ -555,3 +518,41 @@ DB_SHOW_COMMAND(spr, db_show_spr)
(unsigned long)spr);
}
#endif
+
+#undef bzero
+void
+bzero(void *buf, size_t len)
+{
+ caddr_t p;
+
+ p = buf;
+
+ while (((vm_offset_t) p & (sizeof(u_long) - 1)) && len) {
+ *p++ = 0;
+ len--;
+ }
+
+ while (len >= sizeof(u_long) * 8) {
+ *(u_long*) p = 0;
+ *((u_long*) p + 1) = 0;
+ *((u_long*) p + 2) = 0;
+ *((u_long*) p + 3) = 0;
+ len -= sizeof(u_long) * 8;
+ *((u_long*) p + 4) = 0;
+ *((u_long*) p + 5) = 0;
+ *((u_long*) p + 6) = 0;
+ *((u_long*) p + 7) = 0;
+ p += sizeof(u_long) * 8;
+ }
+
+ while (len >= sizeof(u_long)) {
+ *(u_long*) p = 0;
+ len -= sizeof(u_long);
+ p += sizeof(u_long);
+ }
+
+ while (len) {
+ *p++ = 0;
+ len--;
+ }
+}
Modified: head/sys/riscv/riscv/machdep.c
==
--- head/sys/riscv/riscv/machdep.c Sat Sep 9 05:50:47 2017
(r323345)
+++ head/sys/riscv/riscv/machdep.c Sat Sep 9 05:56:04 2017
(r323346)
@@ -151,16 +151,6 @@ cpu_idle_wakeup(int cpu)
return (0);
}
-void
-bzero(void *buf, size_t len)
-{
- uint8_t *p;
-
- p = buf;
- while(len-- > 0)
- *p++ = 0;
-}
-
int
fill_regs(struct thread *td, struct reg *regs)
{
@@ -890,4 +880,15 @@ initriscv(struct riscv_bootparams *rvbp)
riscv_init_interrupts();
early_boot = 0;
+}
+
+#undef bzero
+void
+bzero(void *buf, size_t len)
+{
+ uint8_t *p;
+
+ p = buf;
+ while(len-- > 0)
+ *p++ = 0;
}
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323345 - head/usr.bin/tail/tests
Author: ngie
Date: Sat Sep 9 05:50:47 2017
New Revision: 323345
URL: https://svnweb.freebsd.org/changeset/base/323345
Log:
Check result of seq call and save output via atf_check -o
This ensures that seq outputting to ints will not fail as silently if there's
an error.
Modified:
head/usr.bin/tail/tests/tail_test.sh
Modified: head/usr.bin/tail/tests/tail_test.sh
==
--- head/usr.bin/tail/tests/tail_test.shSat Sep 9 05:42:23 2017
(r323344)
+++ head/usr.bin/tail/tests/tail_test.shSat Sep 9 05:50:47 2017
(r323345)
@@ -224,7 +224,7 @@ broken_pipe_body()
{
atf_expect_fail "Can't seem to get testcase to work in test
environment. Reproduces easily in interactive shell."
- seq -f '%128g' 1 1000 > ints
+ atf_check -o save:ints seq -f '%128g' 1 1000
atf_check -s exit:1 -o ignore -e "inline:tail: stdout" tail -n 856 ints
| awk '{ exit }'
}
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323343 - head/sys/dev/cxgbe
Author: np
Date: Sat Sep 9 05:12:14 2017
New Revision: 323343
URL: https://svnweb.freebsd.org/changeset/base/323343
Log:
cxgbe(4): Fix a couple of problems in the sge_wrq data path.
- start_wrq_wr must not drain the wr_list if there are incomplete_wrs
pending. This can happen when a t4_wrq_tx runs between two
start_wrq_wr.
- commit_wrq_wr must examine the cookie's pidx and ndesc with the
queue's lock held. Otherwise there is a bad race when incomplete WRs
are being completed and commit_wrq_wr for the WR that is ahead in the
queue updates the next incomplete WR's cookie's pidx/ndesc but the
commit_wrq_wr for the second one is using stale values that it read
without the lock.
MFC after:1 week
Sponsored by: Chelsio Communications
Modified:
head/sys/dev/cxgbe/t4_sge.c
Modified: head/sys/dev/cxgbe/t4_sge.c
==
--- head/sys/dev/cxgbe/t4_sge.c Sat Sep 9 03:09:02 2017(r323342)
+++ head/sys/dev/cxgbe/t4_sge.c Sat Sep 9 05:12:14 2017(r323343)
@@ -2354,7 +2354,7 @@ start_wrq_wr(struct sge_wrq *wrq, int len16, struct wr
EQ_LOCK(eq);
- if (!STAILQ_EMPTY(&wrq->wr_list))
+ if (TAILQ_EMPTY(&wrq->incomplete_wrs) && !STAILQ_EMPTY(&wrq->wr_list))
drain_wrq_wr_list(sc, wrq);
if (!STAILQ_EMPTY(&wrq->wr_list)) {
@@ -2408,9 +2408,6 @@ commit_wrq_wr(struct sge_wrq *wrq, void *w, struct wrq
return;
}
- ndesc = cookie->ndesc; /* Can be more than SGE_MAX_WR_NDESC here. */
- pidx = cookie->pidx;
- MPASS(pidx >= 0 && pidx < eq->sidx);
if (__predict_false(w == &wrq->ss[0])) {
int n = (eq->sidx - wrq->ss_pidx) * EQ_ESIZE;
@@ -2422,6 +2419,9 @@ commit_wrq_wr(struct sge_wrq *wrq, void *w, struct wrq
wrq->tx_wrs_direct++;
EQ_LOCK(eq);
+ ndesc = cookie->ndesc; /* Can be more than SGE_MAX_WR_NDESC here. */
+ pidx = cookie->pidx;
+ MPASS(pidx >= 0 && pidx < eq->sidx);
prev = TAILQ_PREV(cookie, wrq_incomplete_wrs, link);
next = TAILQ_NEXT(cookie, link);
if (prev == NULL) {
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r323329 - head/sys/sys
> On Sep 8, 2017, at 13:09, Mateusz Guzik wrote: > > Author: mjg > Date: Fri Sep 8 20:09:14 2017 > New Revision: 323329 > URL: https://svnweb.freebsd.org/changeset/base/323329 > > Log: > Allow __builtin_memset instead of bzero for small buffers of known size > > In particular this eliminates function calls and related register > save/restore > when only few writes would suffice. > > Example speed up can be seen in a fstat microbenchmark on AMD Ryzen cpus, > where > the throughput went up by ~4.5%. > > Thanks to cem@ for benchmarking and reviewing the patch. This change breaks all platforms that use gcc. If it’s not reverted in the next 3 hours, I’ll revert it. Please fix ASAP if possible: https://ci.freebsd.org/job/FreeBSD-head-riscv64-build/3668/console . Thanks, -Ngie signature.asc Description: Message signed with OpenPGP using GPGMail
Re: svn commit: r323342 - head/usr.sbin/manctl
In message <201709090309.v89392bq078...@repo.freebsd.org>, Gordon Tetlow writes : > Author: gordon > Date: Sat Sep 9 03:09:02 2017 > New Revision: 323342 > URL: https://svnweb.freebsd.org/changeset/base/323342 > > Log: > The purge option hasn't been implemented since 1994 when we imported this > code. I think it is safe to say it's not going to be. I'm also working to > de-orbit catman, so remove the reference in the manpage. > > Reviewed by:allanjude > > Modified: > head/usr.sbin/manctl/manctl.8 > head/usr.sbin/manctl/manctl.sh > > Modified: head/usr.sbin/manctl/manctl.8 > = > = > --- head/usr.sbin/manctl/manctl.8 Sat Sep 9 02:28:21 2017(r32334 > 1) > +++ head/usr.sbin/manctl/manctl.8 Sat Sep 9 03:09:02 2017(r32334 > 2) > @@ -33,7 +33,6 @@ > .Nm > .Op Fl compress > .Op Fl uncompress > -.Op Fl purge > .Op Fl help > .Ar path ... > .Sh DESCRIPTION > @@ -50,9 +49,6 @@ Print options and exit. > Compress uncompressed man pages (eliminating .so's). > .It Fl uncompress > Uncompress compressed man pages. > -.It Fl purge > -Purge old formatted man pages (not implemented yet). > .El > .Sh SEE ALSO > -.Xr catman 1 , > .Xr man 1 Don't forget this... Index: manctl.8 === --- manctl.8(revision 323342) +++ manctl.8(working copy) @@ -23,7 +23,7 @@ .\" SUCH DAMAGE. .\" .\" $FreeBSD$ -.Dd January 1, 1996 +.Dd September 8, 2017 .Dt MANCTL 8 .Os .Sh NAME -- Cheers, Cy Schubert FreeBSD UNIX: Web: http://www.FreeBSD.org The need of the many outweighs the greed of the few. ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323342 - head/usr.sbin/manctl
Author: gordon
Date: Sat Sep 9 03:09:02 2017
New Revision: 323342
URL: https://svnweb.freebsd.org/changeset/base/323342
Log:
The purge option hasn't been implemented since 1994 when we imported this
code. I think it is safe to say it's not going to be. I'm also working to
de-orbit catman, so remove the reference in the manpage.
Reviewed by: allanjude
Modified:
head/usr.sbin/manctl/manctl.8
head/usr.sbin/manctl/manctl.sh
Modified: head/usr.sbin/manctl/manctl.8
==
--- head/usr.sbin/manctl/manctl.8 Sat Sep 9 02:28:21 2017
(r323341)
+++ head/usr.sbin/manctl/manctl.8 Sat Sep 9 03:09:02 2017
(r323342)
@@ -33,7 +33,6 @@
.Nm
.Op Fl compress
.Op Fl uncompress
-.Op Fl purge
.Op Fl help
.Ar path ...
.Sh DESCRIPTION
@@ -50,9 +49,6 @@ Print options and exit.
Compress uncompressed man pages (eliminating .so's).
.It Fl uncompress
Uncompress compressed man pages.
-.It Fl purge
-Purge old formatted man pages (not implemented yet).
.El
.Sh SEE ALSO
-.Xr catman 1 ,
.Xr man 1
Modified: head/usr.sbin/manctl/manctl.sh
==
--- head/usr.sbin/manctl/manctl.sh Sat Sep 9 02:28:21 2017
(r323341)
+++ head/usr.sbin/manctl/manctl.sh Sat Sep 9 03:09:02 2017
(r323342)
@@ -39,7 +39,6 @@
# this is now two-pass. If possible, .so's
# are replaced with hard links
# uncompress compressed man pages
-# purge old formatted man pages (not implemented yet)
# Things to watch out for:
# Hard links - careful with g(un)zipping!
# .so's - throw everything through soelim before gzip!
@@ -49,16 +48,6 @@
PATH=/bin:/sbin:/usr/bin:/usr/sbin; export PATH
#
-# purge cat? directories
-#
-do_purge()
-{
- echo "purge $@" 2>&1
- echo "not implemented yet\n" 2>&1
-}
-
-
-#
# Uncompress one page
#
uncompress_page()
@@ -351,8 +340,6 @@ ctl_usage()
{
echo "usage: $1 -compress ... " 1>&2
echo " $1 -uncompress ... " 1>&2
- echo " $1 -purge ... " 1>&2
- echo " $1 -purge expire ... " 1>&2
exit 1
}
@@ -375,6 +362,5 @@ if [ $# -lt 2 ] ; then ctl_usage $0 ; fi ;
case "$1" in
-compress) shift ; do_compress "$@" ;;
-uncompress)shift ; do_uncompress "$@" ;;
- -purge) shift ; do_purge "$@" ;;
*) ctl_usage $0 ;;
esac
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323341 - head/share/man/man5
Author: ian Date: Sat Sep 9 02:28:21 2017 New Revision: 323341 URL: https://svnweb.freebsd.org/changeset/base/323341 Log: Clarify that the size option is in bytes by default, but also accepts the usual SI suffixes. Modified: head/share/man/man5/tmpfs.5 Modified: head/share/man/man5/tmpfs.5 == --- head/share/man/man5/tmpfs.5 Sat Sep 9 02:19:44 2017(r323340) +++ head/share/man/man5/tmpfs.5 Sat Sep 9 02:28:21 2017(r323341) @@ -54,7 +54,7 @@ .\" .\" $FreeBSD$ .\" -.Dd January 20, 2017 +.Dd September 8, 2017 .Dt TMPFS 5 .Os .Sh NAME @@ -131,7 +131,9 @@ the file system size, which can be limited with the .Cm size option. .It Cm size -Specifies the total file system size in bytes. +Specifies the total file system size in bytes, unless suffixed +with one of k, m, g, t, or p, which denote byte, kilobyte, +megabyte, gigabyte, terabyte and petabyte respectively. If zero (the default) or a value larger than SIZE_MAX - PAGE_SIZE is given, the available amount of memory (including main memory and swap space) will be used. ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323340 - head/sys/powerpc/mpc85xx
Author: jhibbits Date: Sat Sep 9 02:19:44 2017 New Revision: 323340 URL: https://svnweb.freebsd.org/changeset/base/323340 Log: Add P5021 and P5040 conditions for LAW count check. P5040/P5021 have the same number of LAWs as P5020. There may be a better way of getting the count from the FDT (fsl,num-laws property on soc/corenet-law or soc/ecm-law), but that's not supported everywhere, so we still need this check for those other cases. Modified: head/sys/powerpc/mpc85xx/mpc85xx.c Modified: head/sys/powerpc/mpc85xx/mpc85xx.c == --- head/sys/powerpc/mpc85xx/mpc85xx.c Sat Sep 9 02:08:22 2017 (r323339) +++ head/sys/powerpc/mpc85xx/mpc85xx.c Sat Sep 9 02:19:44 2017 (r323340) @@ -96,6 +96,10 @@ law_getmax(void) break; case SVR_P5020: case SVR_P5020E: + case SVR_P5021: + case SVR_P5021E: + case SVR_P5040: + case SVR_P5040E: law_max = 32; break; default: ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323339 - head/sys/powerpc/include
Author: jhibbits Date: Sat Sep 9 02:08:22 2017 New Revision: 323339 URL: https://svnweb.freebsd.org/changeset/base/323339 Log: Add some more PVR and SVR defines These processors may not be supported yet, but add them for completion. POWER9 is planned for support. e300 may work (based on 603e core). P5040/P5021 are similar to P5020, so should work as well. One addition is needed for P5040, to support the number of LAWs, and will be a separate commit. Modified: head/sys/powerpc/include/spr.h Modified: head/sys/powerpc/include/spr.h == --- head/sys/powerpc/include/spr.h Sat Sep 9 01:41:01 2017 (r323338) +++ head/sys/powerpc/include/spr.h Sat Sep 9 02:08:22 2017 (r323339) @@ -171,6 +171,7 @@ #define IBMPOWER7PLUS 0x004a #define IBMPOWER8E 0x004b #define IBMPOWER8 0x004d +#define IBMPOWER9 0x004e #define MPC860 0x0050 #define IBMCELLBE 0x0070 #define MPC8240 0x0081 @@ -191,6 +192,10 @@ #define FSL_E500mc 0x8023 #define FSL_E5500 0x8024 #define FSL_E6500 0x8040 +#define FSL_E300C1 0x8083 +#define FSL_E300C2 0x8084 +#define FSL_E300C3 0x8085 +#define FSL_E300C4 0x8086 #defineSPR_EPCR0x133 #define EPCR_EXTGS 0x8000 @@ -733,6 +738,10 @@ #define SVR_P4080E 0x8209 #define SVR_P5020 0x8220 #define SVR_P5020E 0x8228 +#define SVR_P5021 0x8205 +#define SVR_P5021E 0x820d +#define SVR_P5040 0x8204 +#define SVR_P5040E 0x820c #defineSVR_VER(svr)(((svr) >> 16) & 0x) #defineSPR_PID00x030 /* ..8 Process ID Register 0 */ ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323338 - head/sys/geom/eli
Author: cem
Date: Sat Sep 9 01:41:01 2017
New Revision: 323338
URL: https://svnweb.freebsd.org/changeset/base/323338
Log:
Fix information leak in geli(8) integrity mode
In integrity mode, a larger logical sector (e.g., 4096 bytes) spans several
physical sectors (e.g., 512 bytes) on the backing device. Due to hash
overhead, a 4096 byte logical sector takes 8.5625 512-byte physical sectors.
This means that only 288 bytes (256 data + 32 hash) of the last 512 byte
sector are used.
The memory allocation used to store the encrypted data to be written to the
physical sectors comes from malloc(9) and does not use M_ZERO.
Previously, nothing initialized the final physical sector backing each
logical sector, aside from the hash + encrypted data portion. So 224 bytes
of kernel heap memory was leaked to every block :-(.
This patch addresses the issue by initializing the trailing portion of the
physical sector in every logical sector to zeros before use. A much simpler
but higher overhead fix would be to tag the entire allocation M_ZERO.
PR: 222077
Reported by: Maxim Khitrov
Reviewed by: emaste
Security: yes
Sponsored by: Dell EMC Isilon
Differential Revision:https://reviews.freebsd.org/D12272
Modified:
head/sys/geom/eli/g_eli_integrity.c
Modified: head/sys/geom/eli/g_eli_integrity.c
==
--- head/sys/geom/eli/g_eli_integrity.c Sat Sep 9 01:23:30 2017
(r323337)
+++ head/sys/geom/eli/g_eli_integrity.c Sat Sep 9 01:41:01 2017
(r323338)
@@ -463,8 +463,16 @@ g_eli_auth_run(struct g_eli_worker *wr, struct bio *bp
authkey = (u_char *)p; p += G_ELI_AUTH_SECKEYLEN;
data_secsize = sc->sc_data_per_sector;
- if ((i % lsec) == 0)
+ if ((i % lsec) == 0) {
data_secsize = decr_secsize % data_secsize;
+ /*
+* Last encrypted sector of each decrypted sector is
+* only partially filled.
+*/
+ if (bp->bio_cmd == BIO_WRITE)
+ memset(data + sc->sc_alen + data_secsize, 0,
+ encr_secsize - sc->sc_alen - data_secsize);
+ }
if (bp->bio_cmd == BIO_READ) {
/* Remember read HMAC. */
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323337 - stable/10/usr.sbin/bsdinstall/scripts
Author: emaste
Date: Sat Sep 9 01:23:30 2017
New Revision: 323337
URL: https://svnweb.freebsd.org/changeset/base/323337
Log:
MFC r322374: bsdinstall: record DHCP config after obtaining lease
Previously we added an ifconfig_$INTERFACE line to rc.conf for each
unsuccessful DCHP attempt.
PR: 219515
Approved by: re (marius)
Sponsored by: The FreeBSD Foundation
Modified:
stable/10/usr.sbin/bsdinstall/scripts/netconfig_ipv4
Directory Properties:
stable/10/ (props changed)
Modified: stable/10/usr.sbin/bsdinstall/scripts/netconfig_ipv4
==
--- stable/10/usr.sbin/bsdinstall/scripts/netconfig_ipv4Sat Sep 9
00:33:19 2017(r323336)
+++ stable/10/usr.sbin/bsdinstall/scripts/netconfig_ipv4Sat Sep 9
01:23:30 2017(r323337)
@@ -48,8 +48,6 @@ esac
dialog --backtitle 'FreeBSD Installer' --title 'Network Configuration' --yesno
'Would you like to use DHCP to configure this interface?' 0 0
if [ $? -eq $DIALOG_OK ]; then
- echo ifconfig_$INTERFACE=\"${IFCONFIG_PREFIX}DHCP\" >>
$BSDINSTALL_TMPETC/._rc.conf.net
-
if [ ! -z $BSDINSTALL_CONFIGCURRENT ]; then
dialog --backtitle 'FreeBSD Installer' --infobox "Acquiring
DHCP lease..." 0 0
err=$( dhclient $INTERFACE 2>&1 )
@@ -59,6 +57,7 @@ if [ $? -eq $DIALOG_OK ]; then
exec $0 ${INTERFACE} "${IFCONFIG_PREFIX}"
fi
fi
+ echo ifconfig_$INTERFACE=\"${IFCONFIG_PREFIX}DHCP\" >>
$BSDINSTALL_TMPETC/._rc.conf.net
exit 0
fi
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323336 - stable/11/usr.sbin/bsdinstall/scripts
Author: emaste
Date: Sat Sep 9 00:33:19 2017
New Revision: 323336
URL: https://svnweb.freebsd.org/changeset/base/323336
Log:
MFC r322374: bsdinstall: record DHCP config after obtaining lease
Previously we added an ifconfig_$INTERFACE line to rc.conf for each
unsuccessful DCHP attempt.
PR: 219515
Sponsored by: The FreeBSD Foundation
Modified:
stable/11/usr.sbin/bsdinstall/scripts/netconfig_ipv4
Directory Properties:
stable/11/ (props changed)
Modified: stable/11/usr.sbin/bsdinstall/scripts/netconfig_ipv4
==
--- stable/11/usr.sbin/bsdinstall/scripts/netconfig_ipv4Fri Sep 8
22:04:28 2017(r323335)
+++ stable/11/usr.sbin/bsdinstall/scripts/netconfig_ipv4Sat Sep 9
00:33:19 2017(r323336)
@@ -48,8 +48,6 @@ esac
dialog --backtitle 'FreeBSD Installer' --title 'Network Configuration' --yesno
'Would you like to use DHCP to configure this interface?' 0 0
if [ $? -eq $DIALOG_OK ]; then
- echo ifconfig_$INTERFACE=\"${IFCONFIG_PREFIX}DHCP\" >>
$BSDINSTALL_TMPETC/._rc.conf.net
-
if [ ! -z $BSDINSTALL_CONFIGCURRENT ]; then
dialog --backtitle 'FreeBSD Installer' --infobox "Acquiring
DHCP lease..." 0 0
err=$( dhclient $INTERFACE 2>&1 )
@@ -59,6 +57,7 @@ if [ $? -eq $DIALOG_OK ]; then
exec $0 ${INTERFACE} "${IFCONFIG_PREFIX}"
fi
fi
+ echo ifconfig_$INTERFACE=\"${IFCONFIG_PREFIX}DHCP\" >>
$BSDINSTALL_TMPETC/._rc.conf.net
exit 0
fi
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323333 - in stable/10/usr.sbin/pw: . tests
Author: emaste
Date: Fri Sep 8 21:16:23 2017
New Revision: 32
URL: https://svnweb.freebsd.org/changeset/base/32
Log:
MFC r322677: pw usermod: handle empty secondary group lists (-G '')
"pw usermod someuser -G ''" is supposed make sure that someuser
doesn't have any secondary group memberships.
Previouly it was a nop because split_groups() only intitialised
"groups" if at least one group was specified. As a result the
existing secondary group memberships were kept.
PR: 221417
Submitted by: Fabian Keil
Approved by: re (kib)
Obtained from:ElectroBSD
Relnotes: yes
Modified:
stable/10/usr.sbin/pw/pw_user.c
stable/10/usr.sbin/pw/tests/pw_usermod_test.sh
Directory Properties:
stable/10/ (props changed)
Modified: stable/10/usr.sbin/pw/pw_user.c
==
--- stable/10/usr.sbin/pw/pw_user.c Fri Sep 8 21:02:15 2017
(r323332)
+++ stable/10/usr.sbin/pw/pw_user.c Fri Sep 8 21:16:23 2017
(r32)
@@ -1089,10 +1089,10 @@ split_groups(StringList **groups, char *groupsstr)
char *p;
char tok[] = ", \t";
+ if (*groups == NULL)
+ *groups = sl_init();
for (p = strtok(groupsstr, tok); p != NULL; p = strtok(NULL, tok)) {
grp = group_from_name_or_id(p);
- if (*groups == NULL)
- *groups = sl_init();
sl_add(*groups, newstr(grp->gr_name));
}
}
Modified: stable/10/usr.sbin/pw/tests/pw_usermod_test.sh
==
--- stable/10/usr.sbin/pw/tests/pw_usermod_test.sh Fri Sep 8 21:02:15
2017(r323332)
+++ stable/10/usr.sbin/pw/tests/pw_usermod_test.sh Fri Sep 8 21:16:23
2017(r32)
@@ -128,6 +128,9 @@ user_mod_nogroups_body() {
atf_check -s exit:0 ${PW} usermod foo -G test3,test4
atf_check -s exit:0 -o inline:"test3\ntest4\n" \
awk -F\: '$4 == "foo" { print $1 }' ${HOME}/group
+ atf_check -s exit:0 ${PW} usermod foo -G ""
+ atf_check -s exit:0 -o empty \
+ awk -F\: '$4 == "foo" { print $1 }' ${HOME}/group
}
atf_test_case user_mod_rename
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323332 - in stable/10/usr.sbin/pw: . tests
Author: emaste
Date: Fri Sep 8 21:02:15 2017
New Revision: 323332
URL: https://svnweb.freebsd.org/changeset/base/323332
Log:
MFC r322678: pw useradd: Validate the user name before creating the entry
Previouly it was possible to create users with spaces in the name with:
pw useradd -u 1234 -g 1234 -n 'test user'
The "-g 1234" is relevant, without it the name was already rejected
as expected:
[fk@test ~]$ sudo pw useradd -u 1234 -n 'test user'
pw: invalid character ` ' at position 4 in userid/group name
Bug unintentionally found with a salt config without explicit name entry:
test user:
user.present:
- uid: 1234
- gid: 1234
- fullname: Test user
- shell: /usr/local/bin/bash
- home: /home/test
- groups:
- wheel
- salt
"Luckily" salt modules rarely bother with input validation either ...
PR: 221416
Submitted by: Fabian Keil
Approved by: re (kib)
Obtained from:ElectroBSD
Modified:
stable/10/usr.sbin/pw/pw_user.c
stable/10/usr.sbin/pw/tests/pw_useradd_test.sh
Directory Properties:
stable/10/ (props changed)
Modified: stable/10/usr.sbin/pw/pw_user.c
==
--- stable/10/usr.sbin/pw/pw_user.c Fri Sep 8 20:41:49 2017
(r323331)
+++ stable/10/usr.sbin/pw/pw_user.c Fri Sep 8 21:02:15 2017
(r323332)
@@ -1204,7 +1204,7 @@ pw_user_add(int argc, char **argv, char *arg1)
if (arg1[strspn(arg1, "0123456789")] == '\0')
id = pw_checkid(arg1, UID_MAX);
else
- name = arg1;
+ name = pw_checkname(arg1, 0);
}
while ((ch = getopt(argc, argv, args)) != -1) {
@@ -1216,7 +1216,7 @@ pw_user_add(int argc, char **argv, char *arg1)
quiet = true;
break;
case 'n':
- name = optarg;
+ name = pw_checkname(optarg, 0);
break;
case 'u':
userid = optarg;
Modified: stable/10/usr.sbin/pw/tests/pw_useradd_test.sh
==
--- stable/10/usr.sbin/pw/tests/pw_useradd_test.sh Fri Sep 8 20:41:49
2017(r323331)
+++ stable/10/usr.sbin/pw/tests/pw_useradd_test.sh Fri Sep 8 21:02:15
2017(r323332)
@@ -176,6 +176,43 @@ user_add_name_too_long_body() {
${PW} useradd name_very_vert_very_very_very_long
}
+atf_test_case user_add_name_with_spaces
+user_add_name_with_spaces_body() {
+ populate_etc_skel
+ atf_check -s exit:65 -e match:"invalid character" \
+ ${PW} useradd 'test user'
+ atf_check -s exit:1 -o empty grep "^test user:.*" $HOME/master.passwd
+ # Try again with -n which uses a slightly different code path.
+ atf_check -s exit:65 -e match:"invalid character" \
+ ${PW} useradd -n 'test user'
+ atf_check -s exit:1 -o empty grep "^test user:.*" $HOME/master.passwd
+}
+
+atf_test_case user_add_name_with_spaces_and_gid_specified
+user_add_name_with_spaces_and_gid_specified_body() {
+ populate_etc_skel
+ gid=12345
+ user_name="test user"
+ # pw useradd should fail because of the space in the user
+ # name, not because the group doesn't exist.
+ atf_check -s exit:65 -e match:"invalid character" \
+ ${PW} useradd "${user_name}" -g ${gid}
+ atf_check -s exit:1 -o empty grep "^${user_name}:.*" $HOME/master.passwd
+ # Try again with -n which uses a slightly different code path.
+ atf_check -s exit:65 -e match:"invalid character" \
+ ${PW} useradd -n "${user_name}" -g ${gid}
+ atf_check -s exit:1 -o empty grep "^${user_name}:.*" $HOME/master.passwd
+ # Make sure the user isn't added even if the group exists
+ atf_check -s exit:0 ${PW} groupadd blafasel -g ${gid}
+ atf_check -s exit:65 -e match:"invalid character" \
+ ${PW} useradd "${user_name}" -g ${gid}
+ atf_check -s exit:1 -o empty grep "^${user_name}:.*" $HOME/master.passwd
+ # Try again with the -n option.
+ atf_check -s exit:65 -e match:"invalid character" \
+ ${PW} useradd -n "${user_name}" -g ${gid}
+ atf_check -s exit:1 -o empty grep "^${user_name}:.*" $HOME/master.passwd
+}
+
atf_test_case user_add_expiration
user_add_expiration_body() {
populate_etc_skel
@@ -415,6 +452,8 @@ atf_init_test_cases() {
atf_add_test_case user_add_password_expiration_date_month
atf_add_test_case user_add_password_expiration_date_relative
atf_add_test_case user_add_name_too_long
+ atf_add_test_case user_add_name_with_spaces
+ atf_add_test_case user_add_name_with_spaces_and_gid_specified
atf_add_test_case user_add_expirat
svn commit: r323331 - stable/10/sys/cddl/contrib/opensolaris/common/zfs
Author: emaste
Date: Fri Sep 8 20:41:49 2017
New Revision: 323331
URL: https://svnweb.freebsd.org/changeset/base/323331
Log:
MFC r323002: zfs: do not advertise unsupported hash algorithms
illumos 4185 ("add new cryptographic checksums to ZFS: SHA-512, Skein,
Edon-R") was intentionally merged only partially in r289422, without
adding support for skein, sha512 and edonr on FreeBSD.
Support for skein and sha512 was added later on (in head), but none of
these are supported in stable/10. Prior to this commit zfs(8) correctly
rejected these algorithms, but with an error message that claimed
support:
fk@r500 ~ $zfs set checksum=edonr tank
cannot set property for 'tank': 'checksum' must be one of 'on | off |
fletcher2 | fletcher4 | sha256 | sha512 | skein | edonr'
(This commit removes sha512 and skein in addition to edonr from the
merge of head's r323002.)
PR: 204055
Submitted by: Fabian Keil
Approved by: re (kib)
Obtained from:ElectroBSD
Modified:
stable/10/sys/cddl/contrib/opensolaris/common/zfs/zfs_prop.c
Directory Properties:
stable/10/ (props changed)
Modified: stable/10/sys/cddl/contrib/opensolaris/common/zfs/zfs_prop.c
==
--- stable/10/sys/cddl/contrib/opensolaris/common/zfs/zfs_prop.cFri Sep
8 20:20:35 2017(r323330)
+++ stable/10/sys/cddl/contrib/opensolaris/common/zfs/zfs_prop.cFri Sep
8 20:41:49 2017(r323331)
@@ -241,12 +241,12 @@ zfs_prop_init(void)
zprop_register_index(ZFS_PROP_CHECKSUM, "checksum",
ZIO_CHECKSUM_DEFAULT, PROP_INHERIT, ZFS_TYPE_FILESYSTEM |
ZFS_TYPE_VOLUME,
- "on | off | fletcher2 | fletcher4 | sha256 | sha512 | "
- "skein | edonr", "CHECKSUM", checksum_table);
+ "on | off | fletcher2 | fletcher4 | sha256",
+ "CHECKSUM", checksum_table);
zprop_register_index(ZFS_PROP_DEDUP, "dedup", ZIO_CHECKSUM_OFF,
PROP_INHERIT, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME,
- "on | off | verify | sha256[,verify], sha512[,verify], "
- "skein[,verify], edonr,verify", "DEDUP", dedup_table);
+ "on | off | verify | sha256[,verify]",
+ "DEDUP", dedup_table);
zprop_register_index(ZFS_PROP_COMPRESSION, "compression",
ZIO_COMPRESS_DEFAULT, PROP_INHERIT,
ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME,
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323330 - in head/sys/dev: mpr mps
Author: scottl
Date: Fri Sep 8 20:20:35 2017
New Revision: 323330
URL: https://svnweb.freebsd.org/changeset/base/323330
Log:
Refactor interrupt allocation and deallocation. Add some extra
diagnostics. No other functional changes.
Sponsored by: Netflix
Modified:
head/sys/dev/mpr/mpr_pci.c
head/sys/dev/mps/mps_pci.c
Modified: head/sys/dev/mpr/mpr_pci.c
==
--- head/sys/dev/mpr/mpr_pci.c Fri Sep 8 20:09:14 2017(r323329)
+++ head/sys/dev/mpr/mpr_pci.c Fri Sep 8 20:20:35 2017(r323330)
@@ -268,10 +268,21 @@ mpr_pci_alloc_interrupts(struct mpr_softc *sc)
if ((error != 0) && (sc->disable_msi == 0) &&
((msgs = pci_msi_count(dev)) >= MPR_MSI_COUNT))
error = mpr_alloc_msi(sc, MPR_MSI_COUNT);
- if (error != 0)
- msgs = 0;
+ if (error != 0) {
+ /*
+* If neither MSI or MSI-X are available, assume legacy INTx.
+* This also implies that there will be only 1 queue.
+*/
+ sc->mpr_flags |= MPR_FLAGS_INTX;
+ msgs = 1;
+ } else {
+ sc->mpr_flags |= MPR_FLAGS_MSI;
+ msgs = MPR_MSI_COUNT; /* XXX */
+ }
sc->msi_msgs = msgs;
+ mpr_dprint(sc, MPR_INIT, "Allocated %d interrupts\n", msgs);
+
return (error);
}
@@ -279,47 +290,45 @@ int
mpr_pci_setup_interrupts(struct mpr_softc *sc)
{
device_t dev;
- int i, error;
+ void *ihandler;
+ int i, error, rid, initial_rid;
dev = sc->mpr_dev;
error = ENXIO;
- if (sc->msi_msgs == 0) {
- sc->mpr_flags |= MPR_FLAGS_INTX;
- sc->mpr_irq_rid[0] = 0;
- sc->mpr_irq[0] = bus_alloc_resource_any(dev, SYS_RES_IRQ,
- &sc->mpr_irq_rid[0], RF_SHAREABLE | RF_ACTIVE);
- if (sc->mpr_irq[0] == NULL) {
- mpr_printf(sc, "Cannot allocate INTx interrupt\n");
- return (ENXIO);
- }
- error = bus_setup_intr(dev, sc->mpr_irq[0],
- INTR_TYPE_BIO | INTR_MPSAFE, NULL, mpr_intr, sc,
- &sc->mpr_intrhand[0]);
- if (error)
- mpr_printf(sc, "Cannot setup INTx interrupt\n");
+ if (sc->mpr_flags & MPR_FLAGS_INTX) {
+ initial_rid = 0;
+ ihandler = mpr_intr;
+ } else if (sc->mpr_flags & MPR_FLAGS_MSI) {
+ initial_rid = 1;
+ ihandler = mpr_intr_msi;
} else {
- sc->mpr_flags |= MPR_FLAGS_MSI;
- for (i = 0; i < MPR_MSI_COUNT; i++) {
- sc->mpr_irq_rid[i] = i + 1;
- sc->mpr_irq[i] = bus_alloc_resource_any(dev,
- SYS_RES_IRQ, &sc->mpr_irq_rid[i], RF_ACTIVE);
- if (sc->mpr_irq[i] == NULL) {
- mpr_printf(sc,
- "Cannot allocate MSI interrupt\n");
- return (ENXIO);
- }
- error = bus_setup_intr(dev, sc->mpr_irq[i],
- INTR_TYPE_BIO | INTR_MPSAFE, NULL, mpr_intr_msi,
- sc, &sc->mpr_intrhand[i]);
- if (error) {
- mpr_printf(sc,
- "Cannot setup MSI interrupt %d\n", i);
- break;
- }
+ mpr_dprint(sc, MPR_ERROR|MPR_INIT,
+ "Unable to set up interrupts\n");
+ return (EINVAL);
+ }
+
+ for (i = 0; i < sc->msi_msgs; i++) {
+ rid = i + initial_rid;
+ sc->mpr_irq_rid[i] = rid;
+ sc->mpr_irq[i] = bus_alloc_resource_any(dev, SYS_RES_IRQ,
+ &sc->mpr_irq_rid[i], RF_ACTIVE);
+ if (sc->mpr_irq[i] == NULL) {
+ mpr_dprint(sc, MPR_ERROR|MPR_INIT,
+ "Cannot allocate interrupt RID %d\n", rid);
+ break;
}
+ error = bus_setup_intr(dev, sc->mpr_irq[i],
+ INTR_TYPE_BIO | INTR_MPSAFE, NULL, mpr_intr_msi,
+ sc, &sc->mpr_intrhand[i]);
+ if (error) {
+ mpr_dprint(sc, MPR_ERROR|MPR_INIT,
+ "Cannot setup interrupt RID %d\n", rid);
+ break;
+ }
}
+mpr_dprint(sc, MPR_INIT, "Set up %d interrupts\n", sc->msi_msgs);
return (error);
}
@@ -347,24 +356,17 @@ mpr_pci_free(struct mpr_softc *sc)
bus_dma_tag_destroy(sc->mpr_parent_dmat);
}
- if (sc->mpr_flags & MPR_FLAGS_MSI) {
- for (i = 0; i < MPR_MSI_COUNT; i++) {
- if (s
svn commit: r323329 - head/sys/sys
Author: mjg
Date: Fri Sep 8 20:09:14 2017
New Revision: 323329
URL: https://svnweb.freebsd.org/changeset/base/323329
Log:
Allow __builtin_memset instead of bzero for small buffers of known size
In particular this eliminates function calls and related register save/restore
when only few writes would suffice.
Example speed up can be seen in a fstat microbenchmark on AMD Ryzen cpus,
where
the throughput went up by ~4.5%.
Thanks to cem@ for benchmarking and reviewing the patch.
MFC after:1 week
Modified:
head/sys/sys/systm.h
Modified: head/sys/sys/systm.h
==
--- head/sys/sys/systm.hFri Sep 8 20:07:53 2017(r323328)
+++ head/sys/sys/systm.hFri Sep 8 20:09:14 2017(r323329)
@@ -258,6 +258,12 @@ void hexdump(const void *ptr, int length, const char
*
#define ovbcopy(f, t, l) bcopy((f), (t), (l))
void bcopy(const void * _Nonnull from, void * _Nonnull to, size_t len);
void bzero(void * _Nonnull buf, size_t len);
+#define bzero(buf, len) ({ \
+ if (__builtin_constant_p(len) && (len) <= 64) \
+ __builtin_memset((buf), 0, (len)); \
+ else\
+ bzero((buf), (len));\
+})
void explicit_bzero(void * _Nonnull, size_t);
void *memcpy(void * _Nonnull to, const void * _Nonnull from, size_t len);
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323327 - head/sys/x86/pci
Author: kib
Date: Fri Sep 8 19:51:03 2017
New Revision: 323327
URL: https://svnweb.freebsd.org/changeset/base/323327
Log:
Enhance qpi.c to make it usable on all Core-microarchitecture Xeons.
Scan all buses for CSR bus, not stopping on the first failed
match. Scan all slots for function 0 on the found bus, for instance on
IvyBridge the slot 0 is not decoded at all. Since the scan is quite
unsafe, and access to the buses is mostly useful for developers,
enable the csr buses scan with the tunable.
Current qpi.c makes too many assumptions about the uncore
configuration buses location and about slots occupied. Also it
restricts itself only to Nehalem CPUs. It is needed on all Core-based
Xeons. On the 2600 v2 (IvyBridge) machine I have access to, the CSR
buses have numbers 31 (BSP socket) and 63 (second socket), and there
is no functions pci0.31.0.0 or pci0.63.0.0. According to the CPU
datasheet, all devices on the uncore bus occupy slots >= 8.
Practically, the attach to config buses is required for the intel-pcm
pcm-memory.x tool to work, for instance.
Reviewed by: jhb (previous version)
Sponsored by: Mellanox Technologies
MFC after:1 week
Differential revision:https://reviews.freebsd.org/D12268
Modified:
head/sys/x86/pci/qpi.c
Modified: head/sys/x86/pci/qpi.c
==
--- head/sys/x86/pci/qpi.c Fri Sep 8 19:45:37 2017(r323326)
+++ head/sys/x86/pci/qpi.c Fri Sep 8 19:51:03 2017(r323327)
@@ -63,13 +63,19 @@ static MALLOC_DEFINE(M_QPI, "qpidrv", "qpi system devi
static void
qpi_identify(driver_t *driver, device_t parent)
{
+ int do_qpi;
/* Check CPUID to ensure this is an i7 CPU of some sort. */
- if (!(cpu_vendor_id == CPU_VENDOR_INTEL &&
- CPUID_TO_FAMILY(cpu_id) == 0x6 &&
- (CPUID_TO_MODEL(cpu_id) == 0x1a || CPUID_TO_MODEL(cpu_id) == 0x2c)))
+ if (cpu_vendor_id != CPU_VENDOR_INTEL ||
+ CPUID_TO_FAMILY(cpu_id) != 0x6)
return;
+ /* Only discover buses with configuration devices if allowed by user */
+ do_qpi = 0;
+ TUNABLE_INT_FETCH("hw.attach_intel_csr_pci", &do_qpi);
+ if (!do_qpi)
+ return;
+
/* PCI config register access is required. */
if (pci_cfgregopen() == 0)
return;
@@ -97,6 +103,7 @@ qpi_probe_pcib(device_t dev, int bus)
struct qpi_device *qdev;
device_t child;
uint32_t devid;
+ int s;
/*
* If a PCI bus already exists for this bus number, then
@@ -106,18 +113,23 @@ qpi_probe_pcib(device_t dev, int bus)
return (EEXIST);
/*
-* Attempt to read the device id for device 0, function 0 on
-* the bus. A value of 0x means that the bus is not
-* present.
+* Attempt to read the device id for every slot, function 0 on
+* the bus. If all read values are 0x this means that
+* the bus is not present.
*/
- devid = pci_cfgregread(bus, 0, 0, PCIR_DEVVENDOR, 4);
+ for (s = 0; s <= PCI_SLOTMAX; s++) {
+ devid = pci_cfgregread(bus, s, 0, PCIR_DEVVENDOR, 4);
+ if (devid != 0x)
+ break;
+ }
if (devid == 0x)
return (ENOENT);
if ((devid & 0x) != 0x8086) {
- device_printf(dev,
- "Device at pci%d.0.0 has non-Intel vendor 0x%x\n", bus,
- devid & 0x);
+ if (bootverbose)
+ device_printf(dev,
+ "Device at pci%d.%d.0 has non-Intel vendor 0x%x\n",
+ bus, s, devid & 0x);
return (ENXIO);
}
@@ -137,12 +149,12 @@ qpi_attach(device_t dev)
int bus;
/*
-* Each processor socket has a dedicated PCI bus counting down from
-* 255. We keep probing buses until one fails.
+* Each processor socket has a dedicated PCI bus, sometimes
+* not enumerated by ACPI. Probe all unattached buses from 0
+* to 255.
*/
- for (bus = 255;; bus--)
- if (qpi_probe_pcib(dev, bus) != 0)
- break;
+ for (bus = PCI_BUSMAX; bus >= 0; bus--)
+ qpi_probe_pcib(dev, bus);
return (bus_generic_attach(dev));
}
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323326 - head/sys/x86/iommu
Author: kib
Date: Fri Sep 8 19:45:37 2017
New Revision: 323326
URL: https://svnweb.freebsd.org/changeset/base/323326
Log:
Use IOAPIC PCI rid as the interrupt TLP source id for DMAR interrupt
remapping.
VT-d specification requires use of PCI rid as source id for IOAPICs
enumerated by PCI bus. The values from the DMAR ACPI table should be
only used when IOAPIC is not on PCI.
Reviewed by: jhb
Sponsored by: The FreeBSD Foundation
Hardware provided by: Intel
MFC after:2 weeks
Differential revision:https://reviews.freebsd.org/D12205
Modified:
head/sys/x86/iommu/intel_drv.c
Modified: head/sys/x86/iommu/intel_drv.c
==
--- head/sys/x86/iommu/intel_drv.c Fri Sep 8 19:39:20 2017
(r323325)
+++ head/sys/x86/iommu/intel_drv.c Fri Sep 8 19:45:37 2017
(r323326)
@@ -71,6 +71,9 @@ __FBSDID("$FreeBSD$");
#ifdef DEV_APIC
#include "pcib_if.h"
+#include
+#include
+#include
#endif
#defineDMAR_FAULT_IRQ_RID 0
@@ -788,6 +791,9 @@ dmar_find_nonpci(u_int id, u_int entry_type, uint16_t
ACPI_DMAR_DEVICE_SCOPE *devscope;
ACPI_DMAR_PCI_PATH *path;
char *ptr, *ptrend;
+#ifdef DEV_APIC
+ int error;
+#endif
int i;
for (i = 0; i < dmar_devcnt; i++) {
@@ -809,6 +815,17 @@ dmar_find_nonpci(u_int id, u_int entry_type, uint16_t
continue;
if (devscope->EnumerationId != id)
continue;
+#ifdef DEV_APIC
+ if (entry_type == ACPI_DMAR_SCOPE_TYPE_IOAPIC) {
+ error = ioapic_get_rid(id, rid);
+ /*
+* If our IOAPIC has PCI bindings then
+* use the PCI device rid.
+*/
+ if (error == 0)
+ return (unit);
+ }
+#endif
if (devscope->Length - sizeof(ACPI_DMAR_DEVICE_SCOPE)
== 2) {
if (rid != NULL) {
@@ -818,12 +835,11 @@ dmar_find_nonpci(u_int id, u_int entry_type, uint16_t
path->Device, path->Function);
}
return (unit);
- } else {
- /* XXXKIB */
- printf(
- "dmar_find_nonpci: id %d type %d path length != 2\n",
- id, entry_type);
}
+ printf(
+ "dmar_find_nonpci: id %d type %d path length != 2\n",
+ id, entry_type);
+ break;
}
}
return (NULL);
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323325 - in head/sys/x86: include x86
Author: kib
Date: Fri Sep 8 19:39:20 2017
New Revision: 323325
URL: https://svnweb.freebsd.org/changeset/base/323325
Log:
Add an ioapic_get_rid() function to obtain PCIe TLP requester-id for
the interrupt messages from given IOAPIC, if the IOAPIC can be
enumerated on PCI bus.
If IOAPIC has PCI binding, match the PCI device against MADT
enumerated IOAPIC. Match is done first by registers window physical
address, then by IOAPIC ID as read from the APIC ID register.
PCI bsf address of the matched PCI device is the rid.
Reviewed by: jhb
Sponsored by: The FreeBSD Foundation
Hardware provided by: Intel
MFC after:2 weeks
X-Differential revision: https://reviews.freebsd.org/D12205
Modified:
head/sys/x86/include/apicvar.h
head/sys/x86/x86/io_apic.c
Modified: head/sys/x86/include/apicvar.h
==
--- head/sys/x86/include/apicvar.h Fri Sep 8 19:25:11 2017
(r323324)
+++ head/sys/x86/include/apicvar.h Fri Sep 8 19:39:20 2017
(r323325)
@@ -478,6 +478,8 @@ voidlapic_handle_error(void);
void lapic_handle_intr(int vector, struct trapframe *frame);
void lapic_handle_timer(struct trapframe *frame);
+intioapic_get_rid(u_int apic_id, uint16_t *ridp);
+
extern int x2apic_mode;
extern int lapic_eoi_suppression;
Modified: head/sys/x86/x86/io_apic.c
==
--- head/sys/x86/x86/io_apic.c Fri Sep 8 19:25:11 2017(r323324)
+++ head/sys/x86/x86/io_apic.c Fri Sep 8 19:39:20 2017(r323325)
@@ -38,6 +38,7 @@ __FBSDID("$FreeBSD$");
#include
#include
#include
+#include
#include
#include
@@ -99,6 +100,9 @@ struct ioapic {
volatile ioapic_t *io_addr; /* XXX: should use bus_space */
vm_paddr_t io_paddr;
STAILQ_ENTRY(ioapic) io_next;
+ device_t pci_dev; /* matched pci device, if found */
+ struct resource *pci_wnd; /* BAR 0, should be same or alias to
+ io_paddr */
struct ioapic_intsrc io_pins[0];
};
@@ -622,6 +626,8 @@ ioapic_create(vm_paddr_t addr, int32_t apic_id, int in
io = malloc(sizeof(struct ioapic) +
numintr * sizeof(struct ioapic_intsrc), M_IOAPIC, M_WAITOK);
io->io_pic = ioapic_template;
+ io->pci_dev = NULL;
+ io->pci_wnd = NULL;
mtx_lock_spin(&icu_lock);
io->io_id = next_id++;
io->io_apic_id = ioapic_read(apic, IOAPIC_ID) >> APIC_ID_SHIFT;
@@ -954,7 +960,72 @@ ioapic_pci_probe(device_t dev)
static int
ioapic_pci_attach(device_t dev)
{
+ struct resource *res;
+ volatile ioapic_t *apic;
+ struct ioapic *io;
+ int rid;
+ u_int apic_id;
+ /*
+* Try to match the enumerated ioapic. Match BAR start
+* against io_paddr. Due to a fear that PCI window is not the
+* same as the MADT reported io window, but an alias, read the
+* APIC ID from the mapped BAR and match against it.
+*/
+ rid = PCIR_BAR(0);
+ res = bus_alloc_resource_any(dev, SYS_RES_MEMORY, &rid,
+ RF_ACTIVE | RF_SHAREABLE);
+ if (res == NULL) {
+ if (bootverbose)
+ device_printf(dev, "cannot activate BAR0\n");
+ return (ENXIO);
+ }
+ apic = (volatile ioapic_t *)rman_get_virtual(res);
+ if (rman_get_size(res) < IOAPIC_WND_SIZE) {
+ if (bootverbose)
+ device_printf(dev,
+ "BAR0 too small (%jd) for IOAPIC window\n",
+ (uintmax_t)rman_get_size(res));
+ goto fail;
+ }
+ mtx_lock_spin(&icu_lock);
+ apic_id = ioapic_read(apic, IOAPIC_ID) >> APIC_ID_SHIFT;
+ /* First match by io window address */
+ STAILQ_FOREACH(io, &ioapic_list, io_next) {
+ if (io->io_paddr == (vm_paddr_t)rman_get_start(res))
+ goto found;
+ }
+ /* Then by apic id */
+ STAILQ_FOREACH(io, &ioapic_list, io_next) {
+ if (io->io_id == apic_id)
+ goto found;
+ }
+ mtx_unlock_spin(&icu_lock);
+ if (bootverbose)
+ device_printf(dev,
+ "cannot match pci bar apic id %d against MADT\n",
+ apic_id);
+fail:
+ bus_release_resource(dev, SYS_RES_MEMORY, rid, res);
+ return (ENXIO);
+found:
+ KASSERT(io->pci_dev == NULL,
+ ("ioapic %d pci_dev not NULL", io->io_id));
+ KASSERT(io->pci_wnd == NULL,
+ ("ioapic %d pci_wnd not NULL", io->io_id));
+
+ io->pci_dev = dev;
+ io->pci_wnd = res;
+ if (bootverbose && (io->io_paddr != (vm_paddr_t)rman_get_start(res) ||
+ io->io_id != apic_id)) {
+ device_printf(dev, "pci%d:%d:%d:%d pci BAR0@%jx id %d "
+
svn commit: r323324 - head/sys/x86/include
Author: kib Date: Fri Sep 8 19:25:11 2017 New Revision: 323324 URL: https://svnweb.freebsd.org/changeset/base/323324 Log: Add a constant specifying the min size of the IOAPIC registers window. Sponsored by: The FreeBSD Foundation MFC after:1 week Modified: head/sys/x86/include/apicreg.h Modified: head/sys/x86/include/apicreg.h == --- head/sys/x86/include/apicreg.h Fri Sep 8 19:20:42 2017 (r323323) +++ head/sys/x86/include/apicreg.h Fri Sep 8 19:25:11 2017 (r323324) @@ -469,6 +469,8 @@ typedef struct IOAPIC ioapic_t; #define IOAPIC_WINDOW 0x10 #define IOAPIC_EOIR0x40 +#defineIOAPIC_WND_SIZE 0x50 + /* indexes into IO APIC */ #define IOAPIC_ID 0x00 #define IOAPIC_VER 0x01 ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323323 - head/share/mk
Author: bdrewery Date: Fri Sep 8 19:20:42 2017 New Revision: 323323 URL: https://svnweb.freebsd.org/changeset/base/323323 Log: Tweak comment for install -S usage since it does not impact the build. The -S flag is currently ignored for builds since we filter through tools/install.sh that is intended for both non-root and cross-builds. Sponsored by: Dell EMC Isilon X-MFC-With: r322565 Modified: head/share/mk/bsd.lib.mk Modified: head/share/mk/bsd.lib.mk == --- head/share/mk/bsd.lib.mkFri Sep 8 18:32:13 2017(r323322) +++ head/share/mk/bsd.lib.mkFri Sep 8 19:20:42 2017(r323323) @@ -327,9 +327,9 @@ _EXTRADEPEND: SHLINSTALLFLAGS+= -fschg .endif .endif -# Install libraries with -S to avoid linker races with WORLDTMP and risk -# of modifying in-use libraries when installing to a running system. -# It is safe to avoid this for NO_ROOT builds that are only creating an image. +# Install libraries with -S to avoid risk of modifying in-use libraries when +# installing to a running system. It is safe to avoid this for NO_ROOT builds +# that are only creating an image. .if !defined(NO_SAFE_LIBINSTALL) && !defined(NO_ROOT) SHLINSTALLFLAGS+= -S .endif ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323322 - head/sys/compat/freebsd32
Author: sobomax
Date: Fri Sep 8 18:32:13 2017
New Revision: 323322
URL: https://svnweb.freebsd.org/changeset/base/323322
Log:
Correct bintime32 declaration: uint32_t sec -> time32_t sec.
Submitted by: jhb
MFC after:1 month
Modified:
head/sys/compat/freebsd32/freebsd32.h
Modified: head/sys/compat/freebsd32/freebsd32.h
==
--- head/sys/compat/freebsd32/freebsd32.h Fri Sep 8 18:03:34 2017
(r323321)
+++ head/sys/compat/freebsd32/freebsd32.h Fri Sep 8 18:32:13 2017
(r323322)
@@ -79,7 +79,7 @@ struct itimerspec32 {
} while (0)
struct bintime32 {
- uint32_t sec;
+ time32_t sec;
uint32_t frac[2];
};
#define BT_CP(src, dst, fld) do { \
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323321 - head/sys/dev/bnxt
Author: shurd
Date: Fri Sep 8 18:03:34 2017
New Revision: 323321
URL: https://svnweb.freebsd.org/changeset/base/323321
Log:
Added support for displaying HW port stats using sysctl.
This provides port stats (updated once per second) in
dev.bnxt.X.port_stats for PFs. VFs do not have access to the port stats.
Submitted by: Bhargava Chenna Marreddy
Reviewed by: shurd, sbruno
Approved by: sbruno (mentor)
Sponsored by: Broadcom Limited
Differential Revision:https://reviews.freebsd.org/D11914
Modified:
head/sys/dev/bnxt/bnxt.h
head/sys/dev/bnxt/bnxt_hwrm.c
head/sys/dev/bnxt/bnxt_hwrm.h
head/sys/dev/bnxt/bnxt_sysctl.c
head/sys/dev/bnxt/bnxt_sysctl.h
head/sys/dev/bnxt/if_bnxt.c
Modified: head/sys/dev/bnxt/bnxt.h
==
--- head/sys/dev/bnxt/bnxt.hFri Sep 8 17:51:19 2017(r323320)
+++ head/sys/dev/bnxt/bnxt.hFri Sep 8 18:03:34 2017(r323321)
@@ -558,6 +558,7 @@ struct bnxt_softc {
uint8_t max_tc;
struct bnxt_cos_queue q_info[BNXT_MAX_QUEUE];
+ uint64_tadmin_ticks;
struct iflib_dma_info hw_rx_port_stats;
struct iflib_dma_info hw_tx_port_stats;
struct rx_port_stats*rx_port_stats;
Modified: head/sys/dev/bnxt/bnxt_hwrm.c
==
--- head/sys/dev/bnxt/bnxt_hwrm.c Fri Sep 8 17:51:19 2017
(r323320)
+++ head/sys/dev/bnxt/bnxt_hwrm.c Fri Sep 8 18:03:34 2017
(r323321)
@@ -820,6 +820,25 @@ fail:
}
int
+bnxt_hwrm_port_qstats(struct bnxt_softc *softc)
+{
+ struct hwrm_port_qstats_input req = {0};
+ int rc = 0;
+
+ bnxt_hwrm_cmd_hdr_init(softc, &req, HWRM_PORT_QSTATS);
+
+ req.port_id = htole16(softc->pf.port_id);
+ req.rx_stat_host_addr = htole64(softc->hw_rx_port_stats.idi_paddr);
+ req.tx_stat_host_addr = htole64(softc->hw_tx_port_stats.idi_paddr);
+
+ BNXT_HWRM_LOCK(softc);
+ rc = _hwrm_send_message(softc, &req, sizeof(req));
+ BNXT_HWRM_UNLOCK(softc);
+
+ return rc;
+}
+
+int
bnxt_hwrm_cfa_l2_set_rx_mask(struct bnxt_softc *softc,
struct bnxt_vnic_info *vnic)
{
Modified: head/sys/dev/bnxt/bnxt_hwrm.h
==
--- head/sys/dev/bnxt/bnxt_hwrm.h Fri Sep 8 17:51:19 2017
(r323320)
+++ head/sys/dev/bnxt/bnxt_hwrm.h Fri Sep 8 18:03:34 2017
(r323321)
@@ -52,6 +52,7 @@ int bnxt_hwrm_vnic_ctx_alloc(struct bnxt_softc *softc,
int bnxt_hwrm_vnic_cfg(struct bnxt_softc *softc, struct bnxt_vnic_info *vnic);
int bnxt_hwrm_stat_ctx_alloc(struct bnxt_softc *softc, struct bnxt_cp_ring
*cpr,
uint64_t paddr);
+int bnxt_hwrm_port_qstats(struct bnxt_softc *softc);
int bnxt_hwrm_ring_grp_alloc(struct bnxt_softc *softc,
struct bnxt_grp_info *grp);
int bnxt_hwrm_vnic_alloc(struct bnxt_softc *softc, struct bnxt_vnic_info
*vnic);
Modified: head/sys/dev/bnxt/bnxt_sysctl.c
==
--- head/sys/dev/bnxt/bnxt_sysctl.c Fri Sep 8 17:51:19 2017
(r323320)
+++ head/sys/dev/bnxt/bnxt_sysctl.c Fri Sep 8 18:03:34 2017
(r323321)
@@ -164,6 +164,456 @@ bnxt_create_tx_sysctls(struct bnxt_softc *softc, int t
}
int
+bnxt_create_port_stats_sysctls(struct bnxt_softc *softc)
+{
+ struct sysctl_oid *oid;
+ charname[32];
+ chardesc[64];
+
+ sprintf(name, "port_stats");
+ sprintf(desc, "Port Stats");
+ oid = SYSCTL_ADD_NODE(&softc->hw_stats,
+ SYSCTL_CHILDREN(softc->hw_stats_oid), OID_AUTO, name,
CTLFLAG_RD, 0,
+ desc);
+ if (!oid)
+ return ENOMEM;
+
+ SYSCTL_ADD_QUAD(&softc->hw_stats, SYSCTL_CHILDREN(oid), OID_AUTO,
+ "tx_64b_frames", CTLFLAG_RD,
+ &softc->tx_port_stats->tx_64b_frames, "Transmitted 64b frames");
+ SYSCTL_ADD_QUAD(&softc->hw_stats, SYSCTL_CHILDREN(oid), OID_AUTO,
+ "tx_65b_127b_frames", CTLFLAG_RD,
+ &softc->tx_port_stats->tx_65b_127b_frames,
+ "Transmitted 65b 127b frames");
+ SYSCTL_ADD_QUAD(&softc->hw_stats, SYSCTL_CHILDREN(oid), OID_AUTO,
+ "tx_128b_255b_frames", CTLFLAG_RD,
+ &softc->tx_port_stats->tx_128b_255b_frames,
+ "Transmitted 128b 255b frames");
+ SYSCTL_ADD_QUAD(&softc->hw_stats, SYSCTL_CHILDREN(oid), OID_AUTO,
+ "tx_256b_511b_frames", CTLFLAG_RD,
+ &softc->tx_port_stats->tx_256b_511b_frames,
+ "Transmitted 256b 511b frames");
+ SYSCTL_ADD_QUAD(&softc->hw_stats, SYSCTL_CHILDREN(oid), OID_AUTO,
+ "tx_512b_1023b_frames", CTLFLAG_RD,
+ &softc->tx_port_stats->tx_512b_1023b_frames,
+ "Transmitted 512b 1023b frames");
+ SYSCTL_ADD_QUAD(&softc->hw_stats,
svn commit: r323320 - head/sys/dev/mfi
Author: scottl Date: Fri Sep 8 17:51:19 2017 New Revision: 323320 URL: https://svnweb.freebsd.org/changeset/base/323320 Log: Fix intrhook release in MFI as well Modified: head/sys/dev/mfi/mfi.c Modified: head/sys/dev/mfi/mfi.c == --- head/sys/dev/mfi/mfi.c Fri Sep 8 17:40:29 2017(r323319) +++ head/sys/dev/mfi/mfi.c Fri Sep 8 17:51:19 2017(r323320) @@ -1263,8 +1263,6 @@ mfi_startup(void *arg) sc = (struct mfi_softc *)arg; - config_intrhook_disestablish(&sc->mfi_ich); - sc->mfi_enable_intr(sc); sx_xlock(&sc->mfi_config_lock); mtx_lock(&sc->mfi_io_lock); @@ -1273,6 +1271,8 @@ mfi_startup(void *arg) mfi_syspdprobe(sc); mtx_unlock(&sc->mfi_io_lock); sx_xunlock(&sc->mfi_config_lock); + + config_intrhook_disestablish(&sc->mfi_ich); } static void ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323319 - head/sys/dev/amr
Author: scottl
Date: Fri Sep 8 17:40:29 2017
New Revision: 323319
URL: https://svnweb.freebsd.org/changeset/base/323319
Log:
As with r323317, hold off on releasing the intrhook during boot until
we're ready to accept probing from GEOM. Untested, but the pattern is
the same as with aac.
Modified:
head/sys/dev/amr/amr.c
Modified: head/sys/dev/amr/amr.c
==
--- head/sys/dev/amr/amr.c Fri Sep 8 16:59:56 2017(r323318)
+++ head/sys/dev/amr/amr.c Fri Sep 8 17:40:29 2017(r323319)
@@ -302,11 +302,6 @@ amr_startup(void *arg)
debug_called(1);
-/* pull ourselves off the intrhook chain */
-if (sc->amr_ich.ich_func)
- config_intrhook_disestablish(&sc->amr_ich);
-sc->amr_ich.ich_func = NULL;
-
/* get up-to-date drive information */
if (amr_query_controller(sc)) {
device_printf(sc->amr_dev, "can't scan controller for drives\n");
@@ -342,6 +337,11 @@ amr_startup(void *arg)
/* interrupts will be enabled before we do anything more */
sc->amr_state |= AMR_STATE_INTEN;
+
+/* pull ourselves off the intrhook chain */
+if (sc->amr_ich.ich_func)
+ config_intrhook_disestablish(&sc->amr_ich);
+sc->amr_ich.ich_func = NULL;
return;
}
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r323254 - head/sys/compat/freebsd32
On 9/8/17 10:56 AM, Ian Lepore wrote: > On Fri, 2017-09-08 at 08:14 -0400, John Baldwin wrote: >> On 9/7/17 12:29 AM, Maxim Sobolev wrote: >>> >>> Author: sobomax >>> Date: Thu Sep 7 04:29:57 2017 >>> New Revision: 323254 >>> URL: https://svnweb.freebsd.org/changeset/base/323254 >>> >>> Log: >>> In the recvmsg32() system call iterate over returned structure(s) >>> and convert any messages of types SCM_BINTIME, SCM_TIMESTAMP, >>> SCM_REALTIME and SCM_MONOTONIC from 64-bit to its 32-bit >>> representation. Otherwise we either run out of user-supplied >>> buffer to copy those out resulting in the MSG_CTRUNC or simply >>> return values that the userland 32-bit code is not going >>> to parse correctly. This fixes at least two regression tests >>> failing to function properly in 32-bit compat mode: >>> >>> tools/regression/sockets/udp_pingpong >>> tools/regression/sockets/unix_cmsg >>> >>> PR: kern/222039 >>> MFC after:30 days >> Is this correct on !amd64? Other 32-bit platforms use a 64-bit time_t >> (note the time32_t type defined earlier in freebsd32.h). struct bintime32 >> should use time32_t for the seconds field, not uint32_t. I think that >> will be sufficient to make this correct on !amd64 (it also means that >> bintime32 == bintime on !amd64 so you could perhaps use a simpler BT_CP >> for !amd64, but the existing one is probably ok). >> > > The existing one now does *(uint64_t *) on a value that's only aligned > to a 32-bit boundary. That will work in practice because only i386 has > a 32-bit time_t that will use this code, and it's not a strict- > alignment platform. It may still cause compiler warnings about > alignment. Mmmm, the code is used on ppc and mips as well (and mips will fault for unaligned access at least). However, once bintime32 is fixed to use time32_t I think the fraction will be 64-bit aligned? I would be fine with BT_CP() being an #ifdef though that just uses "(dst) = (src)" for !amd64. -- John Baldwin ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323317 - head/sys/dev/aac
Author: scottl Date: Fri Sep 8 16:52:59 2017 New Revision: 323317 URL: https://svnweb.freebsd.org/changeset/base/323317 Log: Move the intrhook release to later in the function so that GEOM knows to wait longer for possible root devices to come online. This fixes a race that seems to be triggered by EARLY_AP_STARTUP. Submitted by: cg...@glup.org Modified: head/sys/dev/aac/aac.c Modified: head/sys/dev/aac/aac.c == --- head/sys/dev/aac/aac.c Fri Sep 8 15:44:52 2017(r323316) +++ head/sys/dev/aac/aac.c Fri Sep 8 16:52:59 2017(r323317) @@ -418,9 +418,6 @@ aac_startup(void *arg) sc = (struct aac_softc *)arg; fwprintf(sc, HBA_FLAGS_DBG_FUNCTION_ENTRY_B, ""); - /* disconnect ourselves from the intrhook chain */ - config_intrhook_disestablish(&sc->aac_ich); - mtx_lock(&sc->aac_io_lock); aac_alloc_sync_fib(sc, &fib); @@ -437,12 +434,15 @@ aac_startup(void *arg) aac_release_sync_fib(sc); mtx_unlock(&sc->aac_io_lock); + /* mark the controller up */ + sc->aac_state &= ~AAC_STATE_SUSPEND; + /* poke the bus to actually attach the child devices */ if (bus_generic_attach(sc->aac_dev)) device_printf(sc->aac_dev, "bus_generic_attach failed\n"); - /* mark the controller up */ - sc->aac_state &= ~AAC_STATE_SUSPEND; + /* disconnect ourselves from the intrhook chain */ + config_intrhook_disestablish(&sc->aac_ich); /* enable interrupts now */ AAC_UNMASK_INTERRUPTS(sc); ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323316 - head/lib/libgeom
Author: cem
Date: Fri Sep 8 15:44:52 2017
New Revision: 323316
URL: https://svnweb.freebsd.org/changeset/base/323316
Log:
libgeom: Remove redundant and duplicated code
In g_open(), g_device_path_open().
No functional change.
Sponsored by: Dell EMC Isilon
Modified:
head/lib/libgeom/geom_util.c
Modified: head/lib/libgeom/geom_util.c
==
--- head/lib/libgeom/geom_util.cFri Sep 8 15:38:02 2017
(r323315)
+++ head/lib/libgeom/geom_util.cFri Sep 8 15:44:52 2017
(r323316)
@@ -56,8 +56,6 @@ g_open(const char *name, int dowrite)
path = g_device_path_open(name, &fd, dowrite);
if (path != NULL)
free(path);
- if (fd == -1)
- return (-1);
return (fd);
}
@@ -281,58 +279,45 @@ g_device_path_open(const char *devpath, int *fdp, int
/* Make sure that we can fail. */
if (fdp != NULL)
*fdp = -1;
+
/* Use the device node if we're able to open it. */
- do {
- fd = open(devpath, dowrite ? O_RDWR : O_RDONLY);
- if (fd == -1)
- break;
- /*
-* Let try to get sectorsize, which will prove it is a GEOM
-* provider.
-*/
- if (g_sectorsize(fd) == -1) {
- close(fd);
- errno = EFTYPE;
- return (NULL);
- }
+ fd = open(devpath, dowrite ? O_RDWR : O_RDONLY);
+ if (fd != -1) {
if ((path = strdup(devpath)) == NULL) {
close(fd);
return (NULL);
}
- if (fdp != NULL)
- *fdp = fd;
- else
- close(fd);
- return (path);
- } while (0);
+ goto fd_ok;
+ }
/* If we're not given an absolute path, assume /dev/ prefix. */
- if (*devpath != '/') {
- asprintf(&path, "%s%s", _PATH_DEV, devpath);
- if (path == NULL)
- return (NULL);
- fd = open(path, dowrite ? O_RDWR : O_RDONLY);
- if (fd == -1) {
- free(path);
- return (NULL);
- }
- /*
-* Let try to get sectorsize, which will prove it is a GEOM
-* provider.
-*/
- if (g_sectorsize(fd) == -1) {
- free(path);
- close(fd);
- errno = EFTYPE;
- return (NULL);
- }
- if (fdp != NULL)
- *fdp = fd;
- else
- close(fd);
- return (path);
+ if (*devpath == '/')
+ return (NULL);
+
+ asprintf(&path, "%s%s", _PATH_DEV, devpath);
+ if (path == NULL)
+ return (NULL);
+ fd = open(path, dowrite ? O_RDWR : O_RDONLY);
+ if (fd == -1) {
+ free(path);
+ return (NULL);
}
- return (NULL);
+
+fd_ok:
+ /*
+* Let try to get sectorsize, which will prove it is a GEOM provider.
+*/
+ if (g_sectorsize(fd) == -1) {
+ free(path);
+ close(fd);
+ errno = EFTYPE;
+ return (NULL);
+ }
+ if (fdp != NULL)
+ *fdp = fd;
+ else
+ close(fd);
+ return (path);
}
char *
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323315 - head/sbin/geom/class/virstor
Author: cem
Date: Fri Sep 8 15:38:02 2017
New Revision: 323315
URL: https://svnweb.freebsd.org/changeset/base/323315
Log:
geom_virstor: Remove wholly unnecessary g_metadata_store copy
Just code cleanup. No functional change.
Sponsored by: Dell EMC Isilon
Modified:
head/sbin/geom/class/virstor/geom_virstor.c
Modified: head/sbin/geom/class/virstor/geom_virstor.c
==
--- head/sbin/geom/class/virstor/geom_virstor.c Fri Sep 8 15:08:17 2017
(r323314)
+++ head/sbin/geom/class/virstor/geom_virstor.c Fri Sep 8 15:38:02 2017
(r323315)
@@ -140,62 +140,6 @@ virstor_main(struct gctl_req *req, unsigned flags)
*/
}
-static void
-pathgen(const char *name, char *path, size_t size)
-{
-
- if (strncmp(name, _PATH_DEV, sizeof(_PATH_DEV) - 1) != 0)
- snprintf(path, size, "%s%s", _PATH_DEV, name);
- else
- strlcpy(path, name, size);
-}
-
-static int
-my_g_metadata_store(const char *name, u_char *md, size_t size)
-{
- char path[MAXPATHLEN];
- unsigned sectorsize;
- off_t mediasize;
- u_char *sector;
- int error, fd;
-
- pathgen(name, path, sizeof(path));
- sector = NULL;
- error = 0;
-
- fd = open(path, O_RDWR);
- if (fd == -1)
- return (errno);
- mediasize = g_get_mediasize(name);
- if (mediasize == 0) {
- error = errno;
- goto out;
- }
- sectorsize = g_get_sectorsize(name);
- if (sectorsize == 0) {
- error = errno;
- goto out;
- }
- assert(sectorsize >= size);
- sector = malloc(sectorsize);
- if (sector == NULL) {
- error = ENOMEM;
- goto out;
- }
- bcopy(md, sector, size);
- bzero(sector + size, sectorsize - size);
- if (pwrite(fd, sector, sectorsize, mediasize - sectorsize) !=
- (ssize_t)sectorsize) {
- error = errno;
- goto out;
- }
-out:
- if (sector != NULL)
- free(sector);
- close(fd);
- return (error);
-}
-
/*
* Labels a new geom Meaning: parses and checks the parameters, calculates &
* writes metadata to the relevant providers so when the next round of
@@ -465,7 +409,7 @@ virstor_label(struct gctl_req *req)
err(1, "Cannot allocate sector of %zu bytes", ssize);
bzero(sect, ssize);
virstor_metadata_encode(&md, sect);
- error = my_g_metadata_store(name, sect, ssize);
+ error = g_metadata_store(name, sect, ssize);
free(sect);
if (error != 0) {
if (verbose)
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r323254 - head/sys/compat/freebsd32
On Fri, 2017-09-08 at 08:14 -0400, John Baldwin wrote: > On 9/7/17 12:29 AM, Maxim Sobolev wrote: > > > > Author: sobomax > > Date: Thu Sep 7 04:29:57 2017 > > New Revision: 323254 > > URL: https://svnweb.freebsd.org/changeset/base/323254 > > > > Log: > > In the recvmsg32() system call iterate over returned structure(s) > > and convert any messages of types SCM_BINTIME, SCM_TIMESTAMP, > > SCM_REALTIME and SCM_MONOTONIC from 64-bit to its 32-bit > > representation. Otherwise we either run out of user-supplied > > buffer to copy those out resulting in the MSG_CTRUNC or simply > > return values that the userland 32-bit code is not going > > to parse correctly. This fixes at least two regression tests > > failing to function properly in 32-bit compat mode: > > > > tools/regression/sockets/udp_pingpong > > tools/regression/sockets/unix_cmsg > > > > PR: kern/222039 > > MFC after:30 days > Is this correct on !amd64? Other 32-bit platforms use a 64-bit time_t > (note the time32_t type defined earlier in freebsd32.h). struct bintime32 > should use time32_t for the seconds field, not uint32_t. I think that > will be sufficient to make this correct on !amd64 (it also means that > bintime32 == bintime on !amd64 so you could perhaps use a simpler BT_CP > for !amd64, but the existing one is probably ok). > The existing one now does *(uint64_t *) on a value that's only aligned to a 32-bit boundary. That will work in practice because only i386 has a 32-bit time_t that will use this code, and it's not a strict- alignment platform. It may still cause compiler warnings about alignment. -- Ian ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323314 - in head/sbin/geom: class/virstor misc
Author: cem
Date: Fri Sep 8 15:08:17 2017
New Revision: 323314
URL: https://svnweb.freebsd.org/changeset/base/323314
Log:
Audit userspace geom code for leaking memory to disk
Any geom class using g_metadata_store, as well as geom_virstor which
duplicated g_metadata_store internally, would dump sectorsize - mdsize bytes
of userspace memory following the metadata block stored. This is most or all
geom classes (gcache, gconcat, geli, gjournal, glabel, gmirror, gmultipath,
graid3, gshsec, gstripe, and geom_virstor).
PR: 222077 (comment #3)
Reported by: Maxim Khitrov
Reviewed by: des
Security: yes
Sponsored by: Dell EMC Isilon
Differential Revision:https://reviews.freebsd.org/D12269
Modified:
head/sbin/geom/class/virstor/geom_virstor.c
head/sbin/geom/misc/subr.c
Modified: head/sbin/geom/class/virstor/geom_virstor.c
==
--- head/sbin/geom/class/virstor/geom_virstor.c Fri Sep 8 14:56:26 2017
(r323313)
+++ head/sbin/geom/class/virstor/geom_virstor.c Fri Sep 8 15:08:17 2017
(r323314)
@@ -183,6 +183,7 @@ my_g_metadata_store(const char *name, u_char *md, size
goto out;
}
bcopy(md, sector, size);
+ bzero(sector + size, sectorsize - size);
if (pwrite(fd, sector, sectorsize, mediasize - sectorsize) !=
(ssize_t)sectorsize) {
error = errno;
Modified: head/sbin/geom/misc/subr.c
==
--- head/sbin/geom/misc/subr.c Fri Sep 8 14:56:26 2017(r323313)
+++ head/sbin/geom/misc/subr.c Fri Sep 8 15:08:17 2017(r323314)
@@ -302,6 +302,7 @@ g_metadata_store(const char *name, const unsigned char
goto out;
}
bcopy(md, sector, size);
+ bzero(sector + size, sectorsize - size);
if (pwrite(fd, sector, sectorsize, mediasize - sectorsize) !=
sectorsize) {
error = errno;
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323312 - head/sys/dev/e1000
Author: kib
Date: Fri Sep 8 14:54:07 2017
New Revision: 323312
URL: https://svnweb.freebsd.org/changeset/base/323312
Log:
Fix malloc() uses in em_get_regs().
Do not use malloc(M_NOWAIT), wait is possible there, and the malloc
failures where not checked. Do not forget to free malloced memory.
Reported and tested by: pho
Approved by: sbruno
Sponsored by: The FreeBSD Foundation
Modified:
head/sys/dev/e1000/if_em.c
Modified: head/sys/dev/e1000/if_em.c
==
--- head/sys/dev/e1000/if_em.c Fri Sep 8 14:35:17 2017(r323311)
+++ head/sys/dev/e1000/if_em.c Fri Sep 8 14:54:07 2017(r323312)
@@ -534,22 +534,26 @@ static int em_get_regs(SYSCTL_HANDLER_ARGS)
{
struct adapter *adapter = (struct adapter *)arg1;
struct e1000_hw *hw = &adapter->hw;
-
struct sbuf *sb;
- u32 *regs_buff = (u32 *)malloc(sizeof(u32) * IGB_REGS_LEN, M_DEVBUF,
M_NOWAIT);
+ u32 *regs_buff;
int rc;
+ regs_buff = malloc(sizeof(u32) * IGB_REGS_LEN, M_DEVBUF, M_WAITOK);
memset(regs_buff, 0, IGB_REGS_LEN * sizeof(u32));
rc = sysctl_wire_old_buffer(req, 0);
MPASS(rc == 0);
- if (rc != 0)
+ if (rc != 0) {
+ free(regs_buff, M_DEVBUF);
return (rc);
+ }
sb = sbuf_new_for_sysctl(NULL, NULL, 32*400, req);
MPASS(sb != NULL);
- if (sb == NULL)
+ if (sb == NULL) {
+ free(regs_buff, M_DEVBUF);
return (ENOMEM);
+ }
/* General Registers */
regs_buff[0] = E1000_READ_REG(hw, E1000_CTRL);
@@ -604,6 +608,8 @@ static int em_get_regs(SYSCTL_HANDLER_ARGS)
sbuf_printf(sb, "\tTDFT\t %08x\n", regs_buff[19]);
sbuf_printf(sb, "\tTDFHS\t %08x\n", regs_buff[20]);
sbuf_printf(sb, "\tTDFPC\t %08x\n\n", regs_buff[21]);
+
+ free(regs_buff, M_DEVBUF);
#ifdef DUMP_DESCS
{
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323311 - svnadmin/conf
Author: pfg Date: Fri Sep 8 14:35:17 2017 New Revision: 323311 URL: https://svnweb.freebsd.org/changeset/base/323311 Log: Add Fedor Uporov (fsu@) as a src committer. He will be having fun with the ext2fs driver and perhaps other slightly related places. Approved by: core Modified: svnadmin/conf/access svnadmin/conf/mentors Modified: svnadmin/conf/access == --- svnadmin/conf/accessFri Sep 8 10:39:28 2017(r323310) +++ svnadmin/conf/accessFri Sep 8 14:35:17 2017(r323311) @@ -73,6 +73,7 @@ eri erj fabient fanf +fsu gabor gad gallatin Modified: svnadmin/conf/mentors == --- svnadmin/conf/mentors Fri Sep 8 10:39:28 2017(r323310) +++ svnadmin/conf/mentors Fri Sep 8 14:35:17 2017(r323311) @@ -16,6 +16,7 @@ arichardson jhb Co-mentor: brooks dabvangyzen defpjd eriae Co-mentor: thompsa +fsupfg gordon delphij Co-mentor: emaste ivadaszadrian Co-mentor: cognet jceel trasz ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r323254 - head/sys/compat/freebsd32
On 9/7/17 12:29 AM, Maxim Sobolev wrote: > Author: sobomax > Date: Thu Sep 7 04:29:57 2017 > New Revision: 323254 > URL: https://svnweb.freebsd.org/changeset/base/323254 > > Log: > In the recvmsg32() system call iterate over returned structure(s) > and convert any messages of types SCM_BINTIME, SCM_TIMESTAMP, > SCM_REALTIME and SCM_MONOTONIC from 64-bit to its 32-bit > representation. Otherwise we either run out of user-supplied > buffer to copy those out resulting in the MSG_CTRUNC or simply > return values that the userland 32-bit code is not going > to parse correctly. This fixes at least two regression tests > failing to function properly in 32-bit compat mode: > > tools/regression/sockets/udp_pingpong > tools/regression/sockets/unix_cmsg > > PR: kern/222039 > MFC after: 30 days Is this correct on !amd64? Other 32-bit platforms use a 64-bit time_t (note the time32_t type defined earlier in freebsd32.h). struct bintime32 should use time32_t for the seconds field, not uint32_t. I think that will be sufficient to make this correct on !amd64 (it also means that bintime32 == bintime on !amd64 so you could perhaps use a simpler BT_CP for !amd64, but the existing one is probably ok). -- John Baldwin ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323310 - head/sys/x86/pci
Author: kib
Date: Fri Sep 8 10:39:28 2017
New Revision: 323310
URL: https://svnweb.freebsd.org/changeset/base/323310
Log:
Consistently use tabs for indent.
Sponsored by: The FreeBSD Foundation
MFC after:1 week
Modified:
head/sys/x86/pci/qpi.c
Modified: head/sys/x86/pci/qpi.c
==
--- head/sys/x86/pci/qpi.c Fri Sep 8 08:02:06 2017(r323309)
+++ head/sys/x86/pci/qpi.c Fri Sep 8 10:39:28 2017(r323310)
@@ -64,15 +64,15 @@ static void
qpi_identify(driver_t *driver, device_t parent)
{
-/* Check CPUID to ensure this is an i7 CPU of some sort. */
-if (!(cpu_vendor_id == CPU_VENDOR_INTEL &&
+ /* Check CPUID to ensure this is an i7 CPU of some sort. */
+ if (!(cpu_vendor_id == CPU_VENDOR_INTEL &&
CPUID_TO_FAMILY(cpu_id) == 0x6 &&
(CPUID_TO_MODEL(cpu_id) == 0x1a || CPUID_TO_MODEL(cpu_id) == 0x2c)))
-return;
+ return;
-/* PCI config register access is required. */
-if (pci_cfgregopen() == 0)
-return;
+ /* PCI config register access is required. */
+ if (pci_cfgregopen() == 0)
+ return;
/* Add a qpi bus device. */
if (BUS_ADD_CHILD(parent, 20, "qpi", -1) == NULL)
@@ -219,7 +219,7 @@ qpi_pcib_attach(device_t dev)
{
device_add_child(dev, "pci", -1);
-return (bus_generic_attach(dev));
+ return (bus_generic_attach(dev));
}
static int
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323309 - head/sys/arm64/arm64
Author: andrew Date: Fri Sep 8 08:02:06 2017 New Revision: 323309 URL: https://svnweb.freebsd.org/changeset/base/323309 Log: Not all CPUs handle reading ID_AA64MMFR2_EL1 (e.g. qemu), disable it for now. Sponsored by: DARPA, AFRL Modified: head/sys/arm64/arm64/identcpu.c Modified: head/sys/arm64/arm64/identcpu.c == --- head/sys/arm64/arm64/identcpu.c Fri Sep 8 06:57:11 2017 (r323308) +++ head/sys/arm64/arm64/identcpu.c Fri Sep 8 08:02:06 2017 (r323309) @@ -897,7 +897,11 @@ identify_cpu(void) cpu_desc[cpu].id_aa64isar1 = READ_SPECIALREG(ID_AA64ISAR1_EL1); cpu_desc[cpu].id_aa64mmfr0 = READ_SPECIALREG(ID_AA64MMFR0_EL1); cpu_desc[cpu].id_aa64mmfr1 = READ_SPECIALREG(ID_AA64MMFR1_EL1); +#ifdef NOTYET cpu_desc[cpu].id_aa64mmfr2 = READ_SPECIALREG(ID_AA64MMFR2_EL1); +#else + cpu_desc[cpu].id_aa64mmfr2 = 0; +#endif cpu_desc[cpu].id_aa64pfr0 = READ_SPECIALREG(ID_AA64PFR0_EL1); cpu_desc[cpu].id_aa64pfr1 = READ_SPECIALREG(ID_AA64PFR1_EL1); ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
