Simon Josefsson wrote:
> > + int alarm_value = 50;
> >signal (SIGALRM, SIG_DFL);
> > - alarm (10);
> > + alarm (alarm_value);
>
> Nice trick, but doesn't the compiler optimize away this?
When I'm compiling for debugging, I use CFLAGS="-ggdb", that is,
implicitly -O0. (Debugging at -O2 is s
Bruno Haible writes:
> + int alarm_value = 50;
>signal (SIGALRM, SIG_DFL);
> - alarm (10);
> + alarm (alarm_value);
Nice trick, but doesn't the compiler optimize away this? Maybe a
'volatile' is needed.
/Simon
signature.asc
Description: PGP signature
With the clang 17 ASAN+UBSAN, I also see this failure:
FAIL: test-u8-strstr
FAIL test-u8-strstr (exit status: 142)
The cause is that the test has a timeout of 5 seconds; however, it needs 7.1
seconds to execute.
This patch increases the timeouts, also leaving some slack for
On x86_64, for which clear_cache is a no-op, I see the 'test-cache' test fail.
How to reproduce:
$ rm -rf ../testdir1; ./gnulib-tool --create-testdir --dir=../testdir1
--single-configure jit/cache
$ cd ../testdir1
Use clang 17 with ASAN and UBSAN:
$ export CC="$clang17dir/bin/clang -Wl,-rpath,$cl
Similarly, the clang 17 ASAN reports the 'test-free' test:
=
==2069723==ERROR: AddressSanitizer: heap-buffer-overflow on address
0x7f8a90324000 at pc 0x55c1bfe66a3a bp 0x7ffdd392c9d0 sp 0x7ffdd392c190
READ of size 4096 at 0x7f8a90324
Building a testdir with clang17 + ASAN + UBSAN, I see these test failures:
FAIL: test-explicit_bzero
FAIL: test-memset_explicit
The sanitizer's complaints are justified, because these tests really do dirty
things with the memory. So, let's skip the relevant parts when ASAN is in use.
2024-05-09
With the clang17 address sanitizer and UB sanitizer, I'm seeing these two
test failures:
FAIL: test-dprintf-posix2.sh
FAIL: test-fprintf-posix3.sh
The cause is that the malloc (1000) call returns NULL. The test cannot
work in this circumstance.
2024-05-09 Bruno Haible
dprintf-po
Building a gnulib testdir of
- some unistdio/*printf modules,
- printf-with-n-directive,
- libunistring-optional,
in an environment where GNU libunistring 1.2 is installed, I see test failures
such as:
FAIL: test-u8-vasnprintf1
=
../../gltests/unistdio/test-u8-printf
Collin Funk wrote:
> I've pushed this patch using stdckdint.h instead of intprops.h in
> ftello. I think this should be the last occurrence of the
> INT_*_OK/INT_*_WRAPV that wasn't yet converted.
Thanks. Looks good.
Bruno
