Paul Eggert wrote:
> Which version of GCC are you using?
s390x-linux-gnu-gcc-10 (Ubuntu 10.3.0-1ubuntu1~20.04) 10.3.0
> I suspect the bug is in the earlier call to timespec_add, as mentioned
> above.
I'll debug that call...
Bruno
On 8/29/21 6:15 PM, Bruno Haible wrote:
../../gltests/test-timespec.c:152: assertion 'eq (timespec_add (a, sumbc),
timespec_add (sum, c))' failed
Aborted (core dumped)
In test-timespec.c:152 the local variables are:
...
ntests = 26
computed_hz = 10
i = 1
a = {tv_sec = -9223372036854775
In a testdir of all of Gnulib, on Linux/s390x, test-timespec fails:
../../gltests/test-timespec.c:152: assertion 'eq (timespec_add (a, sumbc),
timespec_add (sum, c))' failed
Aborted (core dumped)
In test-timespec.c:152 the local variables are:
test = {{tv_sec = -9223372036854775808, tv_nsec = -
* lib/attribute.h (ATTRIBUTE_DEALLOC, ATTRIBUTE_DEALLOC_FREE):
New macros, aliases for the _GL_-prefixed macros added
to gnulib-common.m4 last month.
---
ChangeLog | 7 +++
lib/attribute.h | 8
2 files changed, 15 insertions(+)
diff --git a/ChangeLog b/ChangeLog
index df3f0481f
On Linux/mips, the 'supersede' test fails here:
ASSERT (memcmp (&orig_dev, &statbuf.st_dev, sizeof (dev_t)) == 0);
The reason is that
sizeof (orig_dev) = 8
sizeof (statbuf.st_dev) = 4
sizeof (dev_t) = 8
This patch works around it.
2021-08-29 Bruno Haible
In a QEMU user-mode environment (for various CPUs), the test-execute.sh
and test-spawn-pipe.sh tests fail. This happens with qemu 6.1.0, but did not
happen with qemu 4.1.0 and older versions.
The reason is that when qemu is used to execute a program, the file descriptor
3 is open, referencing /dev
In a QEMU user-mode environment (for various CPUs), the test-posix_spawn-fchdir
test fails.
The reason is that when the program opens "/", it actually is the
/usr/-linux-gnu directory, and the "pwd" program thus prints the string
"/usr/-linux-gnu", not "/" as expected.
This patch skips the parts
In a QEMU user-mode environment (for various CPUs), the test-getcwd.sh test
fails, with exit code 5. This happens with qemu 6.1.0, but did not happen
with qemu 4.1.0 and older versions.
It's not easy to fix, because the code is involved. What I see, is that
* The test invokes rpl_getcwd,
* r
On 8/29/21 11:36 AM, Bruno Haible wrote:
Yes, either an abort() or an integer overflow return indicator would be better
OK, I installed the attached, which takes the integer overflow return route.
>From 6aafd2a92b4bb48937f3e767e51a4b7abf2f2217 Mon Sep 17 00:00:00 2001
From: Paul Eggert
Date:
Paul Eggert wrote:
> > IMO, it would be better to have code like this in base64_encode_alloc:
> >
> > if (inlen < 0)
> > /* This argument is invalid, since the API change from 2021-08-28. */
> > abort ();
>
> Another possibility would be to treat inlen < 0 the same as integer
> overfl
Paul Eggert wrote:
> On 8/29/21 8:55 AM, arn...@skeeve.com wrote:
> > Hi.
> >
> > I had to make the change below to dfa.h to get things to compile
> > in gawk. Please apply this.
>
> Sorry about my typo, and thanks for the fix; I installed it.
No problem. Much thanks.
Arnold
On 8/29/21 3:16 AM, Bruno Haible wrote:
There may be invocations out there, of this
function, with an argument between SIZE_MAX/2 and SIZE_MAX-1. Changing
such calls to be undefined behaviour means that these invocations now
need debugging in the packages that contain them.
Luckily these calls
On 8/29/21 8:55 AM, arn...@skeeve.com wrote:
Hi.
I had to make the change below to dfa.h to get things to compile
in gawk. Please apply this.
Sorry about my typo, and thanks for the fix; I installed it.
The explicit_bzero test (which needs to do things that are outside of the C
standard) fails in different circumstances, due to GCC optimizations:
1) On x86_64, with GCC 11 and option -O3,
as well as on m68k with GCC 10 and option -O3, the compiler creates
two simplified copies of the functio
Hi.
I had to make the change below to dfa.h to get things to compile
in gawk. Please apply this.
Thanks,
Arnold
-
--- /usr/local/src/Gnu/gnulib/lib/dfa.h 2021-08-27 16:50:39.579581132 +0300
+++ support/dfa.h 2021-08-29 18:30:25.101719167 +0300
@@ -50,6
Paul Eggert wrote:
> > Or maybe the cause is the line
> >
> >assume (0 <= inlen);
> >
> > At the entry point of a public function, it is better to use 'assert' than
> > 'assume', IMO. 'assume' means "feel free to crash or press the red button
> > if there is an invalid argument".
>
> 'assume
Hi Simon,
> I'm not sure I understand why idx_t is better than size_t
> here, can you elaborate? Why not ssize_t?
You find a detailed explanation in the comments of idx.h.
> Maybe a compromise is to
> keep the old API but add new APIs with idx_t types and the
> implementation of the old functio
Paul Eggert writes:
> -extern void base64_encode (const char *restrict in, size_t inlen,
> - char *restrict out, size_t outlen);
> +extern void base64_encode (const char *restrict in, idx_t inlen,
> + char *restrict out, idx_t outlen);
Thanks f
On 8/28/21 7:12 AM, Bruno Haible wrote:
Or maybe the cause is the line
assume (0 <= inlen);
At the entry point of a public function, it is better to use 'assert' than
'assume', IMO. 'assume' means "feel free to crash or press the red button
if there is an invalid argument".
'assume' was th
On 8/28/21 5:45 AM, Bruno Haible wrote:
There are two problems with this patch:
1) $ ./gnulib-tool --create-testdir --dir=../testdir5 --single-configure base32
base64
prints messages
gnulib-tool: warning: module base32 depends on a module with an
incompatible license: ialloc
gnuli
20 matches
Mail list logo
