sed-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [platform-testers] new snapshot available: sed-4.3.20-e9ab


From: Jim Meyering
Subject: Re: [platform-testers] new snapshot available: sed-4.3.20-e9ab
Date: Wed, 1 Feb 2017 13:59:12 -0800

On Tue, Jan 31, 2017 at 1:35 PM, Assaf Gordon <address@hidden> wrote:
> Hello,
>
> On Tue, Jan 31, 2017 at 08:53:30PM +0000, Assaf Gordon wrote:
>>
>> On Tue, Jan 31, 2017 at 10:34:52AM -0800, Jim Meyering wrote:
>>>
>>> http://meyering.net/sed/sed-4.3.20-e9ab.tar.xz
>>
>>
>> I encountered possible issue the the dfa-bug when building
>> with musl-libc - will send a separate email.
>
>
> When building sed with musl-libc (./configure CC=musl-gcc),
> the new valgrind test fails (newline-dfa-bug.sh).
>
> I suspect this is a problem with musl/valgrind interaction
> and not a problem in sed/dfa - but other suggestions are welcomed.
>
> ---
>
> The file 'musl-sed-dynamic-DFABUG.valgrind.log' contains
> the error as appears in 'make check' logs.
>
> The file 'musl-sed-dynamic-Q.valgrind.log' contains
> the output of running:
>   valgrind ./sed/sed -Q
> It is the same error as the dfa test file, which leads me to think
> it has nothing to do with sed/dfa.
>
> ---
>
> I've also build sed with musl and static linking.
> Testing the DFA bug results in a very large valgrind report.
> The file 'musl-sed-static-DFABUG.valgrind.log.xz' contains the
> first 1000 errors (warning: it expands to ~600KB).
>
> Running again 'valgrind ./sed/sed -Q' still results
> in valgrind warnings ('musl-sed-static-Q.valgrind.log').

Thank you for all all of that testing.
I have run all of the tests using a valgrind-wrapped sed on Fedora 25,
and there was no failure. Combine that with your seeing the same
failure with -Q, I think we don't have to worry about that musl
failure.



reply via email to

[Prev in Thread] Current Thread [Next in Thread]