On Wed, Jul 25, 2012 at 10:08:10AM +0200, Mark Wielaard wrote:
> Maybe in addition if we really want lots of coverage. But I assumed
> -std=c++0x was more widely used at this time. At the moment the sample
Yeah. E.g. g++ 4.6 knows about -std=c++0x, while it doesn't know
-std=c++11.
Marek
ight pub key on fedorahosted.com), so I'd not be able to push this in case
it's fine. Thanks.
2011-10-04 Marek Polacek
* readelf.c (register_info): Assume the right size of an array.
--- elfutils/src/readelf.c.mp 2011-08-30 10:51:27.869835636 +0200
+++ elfutils/src/readel
On 07/09/2011 12:22 PM, Roland McGrath wrote:
> I haven't tried gcc-4.7, but I've cleaned up the code and I hope it avoids
> those warnings now.
I've just successfully did both `make' and `make check' with gcc4.7. Thanks.
Marek
___
elfutils-dev
)
{
abort ();
}
The resulting coredump is now the `testfile59.bz2.'
To achieve a better coverage I've used both ELF_C_READ{,_MMAP}.
Is it OK?
2011-05-23 Marek Polacek
* dwfl-core-file.c: New test.
* run-dwfl-core-file.sh: Use it.
* testfile59.bz2: New
On 05/19/2011 11:53 AM, Mark Wielaard wrote:
> I created a patch:
> https://bugzilla.redhat.com/show_bug.cgi?id=706040
Thanks for the patch & report.
> With the above patch you can get also src/ coverage. That suggests some
> tests that could be added. For example from inspecting the strip.c
> co
On 05/16/2011 10:00 AM, Mark Wielaard wrote:
> How interesting, I didn't know about --enable-debugpred.
At least something good came out of this wrong patch ;)
> Like Ulrich said, you probably ran the binary in such a way that
> libebl isn't loading the correct backend (in such a case
> ebl_reloc
I've been playing with the debugpred and noticed, that every time when
running the eu-readelf -r, e.g., src/readelf -r src/readelf, we get:
readelf.c:1845: wrong=148, correct=0 < WARNING
If the prediction should be changed, here's a patch.
2011-05-15
Even though this should be handled by the gcc's tree_estimate_probability,
adding the unlikely () to conditions which are almost never true shouldn't
hurt either. Yes, it compiles. OK to apply?
2011-05-14 Marek Polacek
* xmalloc.c (xmalloc, xcalloc, xrealloc): Ad
On 05/12/2011 11:00 AM, Mark Wielaard wrote:
> Looks fine to me.
Applied.
> BTW. How did you test? Could that test be made part of the testsuite?
I've just tried if it even compiles and then used `nm -f sysv'.
Marek
___
elfutils-devel mailing
On 05/11/2011 10:52 AM, Mark Wielaard wrote:
> It seems safe to just remove it.
All right, in this case we also need to remove the redundant
parameters of `show_symbols_sysv'. Tested on x86_64-redhat-linux,
ok to master?
Marek
2011-05-11 Marek Polacek
FWIW, this removes the always true if and also the never executed else branch.
Maybe that unused branch is there for some particular reason, but if not,
I think it can be removed. Should I apply?
2011-05-10 Marek Polacek
* nm.c (show_symbols_sysv): Remove unused if branch
On 05/09/2011 04:12 PM, Roland McGrath wrote:
> That code has had numerous bugs in the past, so tests would be good.
All right, I'll take a look at it, probably next week. Thanks,
Marek
___
elfutils-devel mailing list
elfutils-devel@lists.fedor
On 05/09/2011 01:20 PM, Mark Wielaard wrote:
> I have to agree with roland, that this doesn't really test that much,
> except if you can flip the bits.
Yep, that is true, I've already thrown it away. I just didn't know
how else/better can this be tested.
> How are you measuring coverage?
Using
Hi,
I've tried to check these functions in just one test, since they are very
similar. Thus, I've written the `elf-flag.c' file, which tests them so they
are all 100% covered. Tested on x86_64-redhat-linux. OK?
Marek
2011-05-07 Marek Polacek
* elf-f
On 05/07/2011 07:14 PM, Ulrich Drepper wrote:
> If it compiles it's fine. If not we know where to find you.
:-) Applied.
___
elfutils-devel mailing list
elfutils-devel@lists.fedorahosted.org
https://fedorahosted.org/mailman/listinfo/elfutils-devel
The function `compare_sections_nonrel' most probably should be marked
as static. It is used nowhere else. Tested on x86_64-redhat-linux. OK?
2011-05-07 Marek Polacek
* unstrip.c (compare_sections_nonrel): Mark this function as static.
diff --git a/src/unstrip.c b/src/unst
On 04/27/2011 06:29 PM, Ulrich Drepper wrote:
> This is just creating noise and causing patches people might have not
> to apply. Fix whitespaces when the file gets changed.
All right, not applied. Thanks,
Marek
___
elfutils-devel mailing list
I've found a few whitespaces nits in the *.c and *.h files. I've fixed
them. I intentionally didn't make any ChangeLog entries, since there
are no code changes. Ok?
Signed-off-by: Marek Polacek
---
backends/arm_corenote.c |2 +-
backends/i386_corenote.c |
Applied.
___
elfutils-devel mailing list
elfutils-devel@lists.fedorahosted.org
https://fedorahosted.org/mailman/listinfo/elfutils-devel
ChangeLog files.
Sounds sensible. BTW: Ulrich, would you accept a similar patch for glibc?
I've run
Jim's script and there are quite a lot of duplications. Thanks.
Signed-off-by: Marek Polacek
---
libebl/eblrelativerelocp.c |2 +-
libelf/libelfP.h |2 +-
src/arlib.c
"}'
Here's a patch that fixes all these occurrences. Ok to apply?
Signed-off-by: Marek Polacek
---
libebl/ChangeLog |4
libebl/eblrelativerelocp.c |2 +-
libelf/ChangeLog |4
libelf/libelfP.h |2 +-
src/ChangeLog
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA256
Hi,
I'm trying to add the '-D' option to readelf. This would fulfil this
old request (binutils readelf already has it):
https://bugzilla.redhat.com/show_bug.cgi?id=444621
It is quite clear that we have to first search for the PT_DYNAMIC segment a
On 03/31/2011 08:41 PM, Roland McGrath wrote:
> IIRC elf_clone is largely broken. There was some old posting here where I
> asked about it and never got any answers. Let's get the theory figured out
> before we start adding tests and calling what they do "correct".
Aha, of course. I didn't know
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA256
Here's an attempt to cover elf_clone. I've tried to memcmp elf and cloned
elf, but this way it didn't work so I've used just manual comparison of
values. Tested on x86_64. Ok?
Signed-off-by: Marek Polacek
- ---
tests/Cha
On 03/29/2011 03:15 PM, Mark Wielaard wrote:
> You did forget to mention the other changes in Makefile.am in your
> ChangeLog entry though.
You're right. I've amended that commit. Should I push it to trunk now?
Thanks,
Marek
___
elfuti
uot;synthetically" instead of depending on compiler
> versions/arch.
All right. Should be done now. Pushed to my branch.
Signed-off-by: Marek Polacek
---
tests/ChangeLog |8 ++-
tests/Makefile.am |8 +-
tests/alldts.c | 275 +
On 03/25/2011 04:27 PM, Marek Polacek wrote:
> Ok, another try. I've constructed an ELF file with all the
> DT_* flags set, it is named testfile58.
FWIW, here is source code of program which constructs an ELF
file like this:
/* Create an ELF file with all the DT_* flags set.
C
depends on printf ("%s", NULL) printing
"(null)" because this way it is a way simpler.
I've also fixed date in the last ChangeLog entry. All is available
in branch mpolacek/ebl_dynamic_tag_name. Tested on x86_64. The
patch is also attached. Ok?
Marek
Signe
On 03/22/2011 09:51 AM, Petr Machata wrote:
>> -case DT_PLTREL:
>> - puts (ebl_dynamic_tag_name (ebl, dyn->d_un.d_val, NULL, 0));
>> +case DT_PLTREL:;
>
> ...that semicolon. ^^^
It's necessary, otherwise we'd get an error. After label there has
to be a statement and declaration isn'
Sometimes, e.g. when malformed ELF file, the ebl_dynamic_tag_name()
returns NULL which causes puts() to segfault immediately without
ceremony. Fixed thusly. Tested on x86_64. Ok for trunk?
Signed-off-by: Marek Polacek
---
src/ChangeLog |4
src/readelf.c |6 --
2 files
In libebl/ebldynamictagname.c there should probably be "TLSDESC_GOT" instead of
"TLSDESC_DOT". In elf.h we have #defined DT_TLSDESC_GOT. Patch is available in
branch mpolacek/tlsdesc. Tested on x86_64, ok for trunk?
Marek
___
elfutils-devel ma
On 03/15/2011 05:23 PM, Roland McGrath wrote:
>> This looks fine to me. But note that libebl isn't a stable interface. So
>> I wouldn't spend too much time adding specific tests for it, unless
>> higher level functions really depend on it. So in this case, why not
>> (also) actually read the dyna
I've just pushed another test for function ebl_dynamic_tag_name. This
function is used mainly in elflint, but also in readelf. Test is quite
simple--just print all the possible strings the function can return. Now
the ebl_dynamic_tag_name is fully covered.
I've also fixed last ChangeLog entry
On Wed, 2011-03-02 at 12:33 -0800, Roland McGrath wrote:
> Really the best thing to do is to join the "gitelfutils" group via
> the https://admin.fedoraproject.org/accounts system. Once you're
> approved there, you can push to git directly. The convention is
> that you are free to push any branch
sufficient? Any comments or suggestions are very welcome. Thanks,
Marek
tests/ChangeLog:
2011-03-02 Marek Polacek
* dwarf-getstring.c: New test.
* run-dwarf-getstring.sh: And its wrapper.
* Makefile.am (EXTRA_DIST): Add and update all.
diff -d -urpN src/elfu
35 matches
Mail list logo
