c/mirror_zfs
1
0
Fork 0
mirror of https://git.proxmox.com/git/mirror_zfs.git synced 2024-11-17 10:01:01 +03:00
Code Issues Packages Projects Releases Wiki Activity
dec697ad68
mirror_zfs/lib/libspl/assert.c

94 lines
2.5 KiB
C
Raw Normal View History

Move libspl_assertf into .c file Variadic functions cannot be inlined. libspl_assertf ends up being duplicated in every file that uses it. Fix this by moving the function into a new assert.c. Also move the definition of aok into the new file instead of zone.c. Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Arvind Sankar <nivedita@alum.mit.edu> Closes #10538
2020-07-05 01:24:13 +03:00
/*
* CDDL HEADER START
*
* The contents of this file are subject to the terms of the
* Common Development and Distribution License (the "License").
* You may not use this file except in compliance with the License.
*
* You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
Replace dead opensolaris.org license link The commit replaces all findings of the link: http://www.opensolaris.org/os/licensing with this one: https://opensource.org/licenses/CDDL-1.0 Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Tino Reichardt <milky-zfs@mcmilk.de> Closes #13619
2022-07-12 00:16:13 +03:00
* or https://opensource.org/licenses/CDDL-1.0.
Move libspl_assertf into .c file Variadic functions cannot be inlined. libspl_assertf ends up being duplicated in every file that uses it. Fix this by moving the function into a new assert.c. Also move the definition of aok into the new file instead of zone.c. Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Arvind Sankar <nivedita@alum.mit.edu> Closes #10538
2020-07-05 01:24:13 +03:00
* See the License for the specific language governing permissions
* and limitations under the License.
*
* When distributing Covered Code, include this CDDL HEADER in each
* file and include the License file at usr/src/OPENSOLARIS.LICENSE.
* If applicable, add the following below this CDDL HEADER, with the
* fields enclosed by brackets "[]" replaced with your own identifying
* information: Portions Copyright [yyyy] [name of copyright owner]
*
* CDDL HEADER END
*/
/*
* Copyright 2008 Sun Microsystems, Inc. All rights reserved.
* Use is subject to license terms.
*/
libspl/assert: show process/task details in assert output Makes it much easier to see what thing complained. Getting thread id, program name and thread name vary wildly between Linux and FreeBSD, so those are set up in macros. pthread_getname_np() did not appear in musl until very recently, but the same info has always been available via prctl(PR_GET_NAME), so we use that instead. Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Rob Norris <robn@despairlabs.com> Sponsored-by: https://despairlabs.com/sponsor/ Closes #16140
2024-04-21 14:43:53 +03:00
/*
* Copyright (c) 2024, Rob Norris <robn@despairlabs.com>
*/
Move libspl_assertf into .c file Variadic functions cannot be inlined. libspl_assertf ends up being duplicated in every file that uses it. Fix this by moving the function into a new assert.c. Also move the definition of aok into the new file instead of zone.c. Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Arvind Sankar <nivedita@alum.mit.edu> Closes #10538
2020-07-05 01:24:13 +03:00
#include <assert.h>
libspl/assert: add lock around assertion output If multiple threads trip an assertion at the same moment (quite common), they can be printing at the same time, and their output gets messy. This adds a simple lock around the whole thing, to prevent a second task printing assert output before the first has finished. Additionally, if libspl_assert_ok is not set, abort() is called without dropping the lock, so that any other asserting tasks will be killed before starting any output, rather than only getting part-way through. This is a tradeoff; it's assumed that multiple threads asserting at the same moment are likely the same fault in different instances of a thread, and so there won't be any more useful information from the other tasks anyway. Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Rob Norris <robn@despairlabs.com> Sponsored-by: https://despairlabs.com/sponsor/ Closes #16140
2024-04-28 05:49:58 +03:00
#include <pthread.h>
Move libspl_assertf into .c file Variadic functions cannot be inlined. libspl_assertf ends up being duplicated in every file that uses it. Fix this by moving the function into a new assert.c. Also move the definition of aok into the new file instead of zone.c. Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Arvind Sankar <nivedita@alum.mit.edu> Closes #10538
2020-07-05 01:24:13 +03:00
libspl/assert: show process/task details in assert output Makes it much easier to see what thing complained. Getting thread id, program name and thread name vary wildly between Linux and FreeBSD, so those are set up in macros. pthread_getname_np() did not appear in musl until very recently, but the same info has always been available via prctl(PR_GET_NAME), so we use that instead. Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Rob Norris <robn@despairlabs.com> Sponsored-by: https://despairlabs.com/sponsor/ Closes #16140
2024-04-21 14:43:53 +03:00
#if defined(__linux__)
#include <errno.h>
#include <sys/prctl.h>
#ifdef HAVE_GETTID
#define libspl_gettid() gettid()
#else
#include <sys/syscall.h>
#define libspl_gettid() ((pid_t)syscall(__NR_gettid))
#endif
#define libspl_getprogname() (program_invocation_short_name)
#define libspl_getthreadname(buf, len) \
prctl(PR_GET_NAME, (unsigned long)(buf), 0, 0, 0)
#elif defined(__FreeBSD__)
#include <pthread_np.h>
#define libspl_gettid() pthread_getthreadid_np()
#define libspl_getprogname() getprogname()
#define libspl_getthreadname(buf, len) \
pthread_getname_np(pthread_self(), buf, len);
#endif
Add `--enable-asan` and `--enable-ubsan` switches `configure` now accepts `--enable-asan` and `--enable-ubsan` switches which results in passing `-fsanitize=address` and `-fsanitize=undefined`, respectively, to the compiler. Those flags are enabled in GitHub workflows for ZTS and zloop. Errors reported by both instrumentations are corrected, except for: - Memory leak reporting is (temporarily) suppressed. The cost of fixing them is relatively high compared to the gains. - Checksum computing functions in `module/zcommon/zfs_fletcher*` have UBSan errors suppressed. It is completely impractical to enforce 64-byte payload alignment there due to performance impact. - There's no ASan heap poisoning in `module/zstd/lib/zstd.c`. A custom memory allocator is used there rendering that measure unfeasible. - Memory leaks detection has to be suppressed for `cmd/zvol_id`. `zvol_id` is run by udev with the help of `ptrace(2)`. Tracing is incompatible with memory leaks detection. Reviewed-by: Ahelenia Ziemiańska <nabijaczleweli@nabijaczleweli.xyz> Reviewed-by: George Melikov <mail@gmelikov.ru> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: szubersk <szuberskidamian@gmail.com> Closes #12928
2022-02-04 01:35:38 +03:00
static boolean_t libspl_assert_ok = B_FALSE;
void
libspl_set_assert_ok(boolean_t val)
{
libspl_assert_ok = val;
}
Move libspl_assertf into .c file Variadic functions cannot be inlined. libspl_assertf ends up being duplicated in every file that uses it. Fix this by moving the function into a new assert.c. Also move the definition of aok into the new file instead of zone.c. Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Arvind Sankar <nivedita@alum.mit.edu> Closes #10538
2020-07-05 01:24:13 +03:00
libspl/assert: add lock around assertion output If multiple threads trip an assertion at the same moment (quite common), they can be printing at the same time, and their output gets messy. This adds a simple lock around the whole thing, to prevent a second task printing assert output before the first has finished. Additionally, if libspl_assert_ok is not set, abort() is called without dropping the lock, so that any other asserting tasks will be killed before starting any output, rather than only getting part-way through. This is a tradeoff; it's assumed that multiple threads asserting at the same moment are likely the same fault in different instances of a thread, and so there won't be any more useful information from the other tasks anyway. Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Rob Norris <robn@despairlabs.com> Sponsored-by: https://despairlabs.com/sponsor/ Closes #16140
2024-04-28 05:49:58 +03:00
static pthread_mutex_t assert_lock = PTHREAD_MUTEX_INITIALIZER;
Move libspl_assertf into .c file Variadic functions cannot be inlined. libspl_assertf ends up being duplicated in every file that uses it. Fix this by moving the function into a new assert.c. Also move the definition of aok into the new file instead of zone.c. Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Arvind Sankar <nivedita@alum.mit.edu> Closes #10538
2020-07-05 01:24:13 +03:00
/* printf version of libspl_assert */
void
libspl_assertf(const char *file, const char *func, int line,
const char *format, ...)
{
libspl/assert: add lock around assertion output If multiple threads trip an assertion at the same moment (quite common), they can be printing at the same time, and their output gets messy. This adds a simple lock around the whole thing, to prevent a second task printing assert output before the first has finished. Additionally, if libspl_assert_ok is not set, abort() is called without dropping the lock, so that any other asserting tasks will be killed before starting any output, rather than only getting part-way through. This is a tradeoff; it's assumed that multiple threads asserting at the same moment are likely the same fault in different instances of a thread, and so there won't be any more useful information from the other tasks anyway. Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Rob Norris <robn@despairlabs.com> Sponsored-by: https://despairlabs.com/sponsor/ Closes #16140
2024-04-28 05:49:58 +03:00
pthread_mutex_lock(&assert_lock);
Move libspl_assertf into .c file Variadic functions cannot be inlined. libspl_assertf ends up being duplicated in every file that uses it. Fix this by moving the function into a new assert.c. Also move the definition of aok into the new file instead of zone.c. Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Arvind Sankar <nivedita@alum.mit.edu> Closes #10538
2020-07-05 01:24:13 +03:00
va_list args;
libspl/assert: show process/task details in assert output Makes it much easier to see what thing complained. Getting thread id, program name and thread name vary wildly between Linux and FreeBSD, so those are set up in macros. pthread_getname_np() did not appear in musl until very recently, but the same info has always been available via prctl(PR_GET_NAME), so we use that instead. Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Rob Norris <robn@despairlabs.com> Sponsored-by: https://despairlabs.com/sponsor/ Closes #16140
2024-04-21 14:43:53 +03:00
char tname[64];
libspl_getthreadname(tname, sizeof (tname));
fprintf(stderr, "ASSERT at %s:%d:%s()\n", file, line, func);
Move libspl_assertf into .c file Variadic functions cannot be inlined. libspl_assertf ends up being duplicated in every file that uses it. Fix this by moving the function into a new assert.c. Also move the definition of aok into the new file instead of zone.c. Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Arvind Sankar <nivedita@alum.mit.edu> Closes #10538
2020-07-05 01:24:13 +03:00
va_start(args, format);
vfprintf(stderr, format, args);
va_end(args);
Reduce false positives from Static Analyzers Both Clang's Static Analyzer and Synopsys' Coverity would ignore assertions. Following Clang's advice, we annotate our assertions: https://clang-analyzer.llvm.org/annotations.html#custom_assertions This makes both Clang's Static Analyzer and Coverity properly identify assertions. This change reduced Clang's reported defects from 246 to 180. It also reduced the false positives reported by Coverityi by 10, while enabling Coverity to find 9 more defects that previously were false negatives. A couple examples of this would be CID-1524417 and CID-1524423. After submitting a build to coverity with the modified assertions, CID-1524417 disappeared while the report for CID-1524423 no longer claimed that the assertion tripped. Coincidentally, it turns out that it is possible to more accurately annotate our headers than the Coverity modelling file permits in the case of format strings. Since we can do that and this patch annotates headers whenever `__coverity_panic__()` would have been used in the model file, we drop all models that use `__coverity_panic__()` from the model file. Upon seeing the success in eliminating false positives involving assertions, it occurred to me that we could also modify our headers to eliminate coverity's false positives involving byte swaps. We now have coverity specific byteswap macros, that do nothing, to disable Coverity's false positives when we do byte swaps. This allowed us to also drop the byteswap definitions from the model file. Lastly, a model file update has been done beyond the mentioned deletions: * The definitions of `umem_alloc_aligned()`, `umem_alloc()` andi `umem_zalloc()` were originally implemented in a way that was intended to inform coverity that when KM_SLEEP has been passed these functions, they do not return NULL. A small error in how this was done was found, so we correct it. * Definitions for umem_cache_alloc() and umem_cache_free() have been added. In practice, no false positives were avoided by making these changes, but in the interest of correctness from future coverity builds, we make them anyway. Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Ryan Moeller <ryan@iXsystems.com> Signed-off-by: Richard Yao <richard.yao@alumni.stonybrook.edu> Closes #13902
2022-10-01 01:30:12 +03:00
libspl/assert: show process/task details in assert output Makes it much easier to see what thing complained. Getting thread id, program name and thread name vary wildly between Linux and FreeBSD, so those are set up in macros. pthread_getname_np() did not appear in musl until very recently, but the same info has always been available via prctl(PR_GET_NAME), so we use that instead. Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Rob Norris <robn@despairlabs.com> Sponsored-by: https://despairlabs.com/sponsor/ Closes #16140
2024-04-21 14:43:53 +03:00
fprintf(stderr, "\n"
" PID: %-8u COMM: %s\n"
" TID: %-8u NAME: %s\n",
getpid(), libspl_getprogname(),
libspl_gettid(), tname);
Reduce false positives from Static Analyzers Both Clang's Static Analyzer and Synopsys' Coverity would ignore assertions. Following Clang's advice, we annotate our assertions: https://clang-analyzer.llvm.org/annotations.html#custom_assertions This makes both Clang's Static Analyzer and Coverity properly identify assertions. This change reduced Clang's reported defects from 246 to 180. It also reduced the false positives reported by Coverityi by 10, while enabling Coverity to find 9 more defects that previously were false negatives. A couple examples of this would be CID-1524417 and CID-1524423. After submitting a build to coverity with the modified assertions, CID-1524417 disappeared while the report for CID-1524423 no longer claimed that the assertion tripped. Coincidentally, it turns out that it is possible to more accurately annotate our headers than the Coverity modelling file permits in the case of format strings. Since we can do that and this patch annotates headers whenever `__coverity_panic__()` would have been used in the model file, we drop all models that use `__coverity_panic__()` from the model file. Upon seeing the success in eliminating false positives involving assertions, it occurred to me that we could also modify our headers to eliminate coverity's false positives involving byte swaps. We now have coverity specific byteswap macros, that do nothing, to disable Coverity's false positives when we do byte swaps. This allowed us to also drop the byteswap definitions from the model file. Lastly, a model file update has been done beyond the mentioned deletions: * The definitions of `umem_alloc_aligned()`, `umem_alloc()` andi `umem_zalloc()` were originally implemented in a way that was intended to inform coverity that when KM_SLEEP has been passed these functions, they do not return NULL. A small error in how this was done was found, so we correct it. * Definitions for umem_cache_alloc() and umem_cache_free() have been added. In practice, no false positives were avoided by making these changes, but in the interest of correctness from future coverity builds, we make them anyway. Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Ryan Moeller <ryan@iXsystems.com> Signed-off-by: Richard Yao <richard.yao@alumni.stonybrook.edu> Closes #13902
2022-10-01 01:30:12 +03:00
#if !__has_feature(attribute_analyzer_noreturn) && !defined(__COVERITY__)
libspl: staticify buf and pagesize, rename aok to libspl_assert_ok Exporting names this short can easily cause nasty collisions with user code. Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: John Kennedy <john.kennedy@delphix.com> Signed-off-by: Ahelenia Ziemiańska <nabijaczleweli@nabijaczleweli.xyz> Closes #12050
2021-06-03 20:04:13 +03:00
if (libspl_assert_ok) {
libspl/assert: add lock around assertion output If multiple threads trip an assertion at the same moment (quite common), they can be printing at the same time, and their output gets messy. This adds a simple lock around the whole thing, to prevent a second task printing assert output before the first has finished. Additionally, if libspl_assert_ok is not set, abort() is called without dropping the lock, so that any other asserting tasks will be killed before starting any output, rather than only getting part-way through. This is a tradeoff; it's assumed that multiple threads asserting at the same moment are likely the same fault in different instances of a thread, and so there won't be any more useful information from the other tasks anyway. Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Rob Norris <robn@despairlabs.com> Sponsored-by: https://despairlabs.com/sponsor/ Closes #16140
2024-04-28 05:49:58 +03:00
pthread_mutex_unlock(&assert_lock);
Move libspl_assertf into .c file Variadic functions cannot be inlined. libspl_assertf ends up being duplicated in every file that uses it. Fix this by moving the function into a new assert.c. Also move the definition of aok into the new file instead of zone.c. Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Arvind Sankar <nivedita@alum.mit.edu> Closes #10538
2020-07-05 01:24:13 +03:00
return;
}
Reduce false positives from Static Analyzers Both Clang's Static Analyzer and Synopsys' Coverity would ignore assertions. Following Clang's advice, we annotate our assertions: https://clang-analyzer.llvm.org/annotations.html#custom_assertions This makes both Clang's Static Analyzer and Coverity properly identify assertions. This change reduced Clang's reported defects from 246 to 180. It also reduced the false positives reported by Coverityi by 10, while enabling Coverity to find 9 more defects that previously were false negatives. A couple examples of this would be CID-1524417 and CID-1524423. After submitting a build to coverity with the modified assertions, CID-1524417 disappeared while the report for CID-1524423 no longer claimed that the assertion tripped. Coincidentally, it turns out that it is possible to more accurately annotate our headers than the Coverity modelling file permits in the case of format strings. Since we can do that and this patch annotates headers whenever `__coverity_panic__()` would have been used in the model file, we drop all models that use `__coverity_panic__()` from the model file. Upon seeing the success in eliminating false positives involving assertions, it occurred to me that we could also modify our headers to eliminate coverity's false positives involving byte swaps. We now have coverity specific byteswap macros, that do nothing, to disable Coverity's false positives when we do byte swaps. This allowed us to also drop the byteswap definitions from the model file. Lastly, a model file update has been done beyond the mentioned deletions: * The definitions of `umem_alloc_aligned()`, `umem_alloc()` andi `umem_zalloc()` were originally implemented in a way that was intended to inform coverity that when KM_SLEEP has been passed these functions, they do not return NULL. A small error in how this was done was found, so we correct it. * Definitions for umem_cache_alloc() and umem_cache_free() have been added. In practice, no false positives were avoided by making these changes, but in the interest of correctness from future coverity builds, we make them anyway. Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Ryan Moeller <ryan@iXsystems.com> Signed-off-by: Richard Yao <richard.yao@alumni.stonybrook.edu> Closes #13902
2022-10-01 01:30:12 +03:00
#endif
Move libspl_assertf into .c file Variadic functions cannot be inlined. libspl_assertf ends up being duplicated in every file that uses it. Fix this by moving the function into a new assert.c. Also move the definition of aok into the new file instead of zone.c. Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Arvind Sankar <nivedita@alum.mit.edu> Closes #10538
2020-07-05 01:24:13 +03:00
abort();
}
Reference in New Issue Copy Permalink