This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [PING][PATCH][BZ #15022] Correct global-scope dlopen issues in static executables
- From: "Carlos O'Donell" <carlos at redhat dot com>
- To: "Maciej W. Rozycki" <macro at codesourcery dot com>, Roland McGrath <roland at hack dot frob dot com>
- Cc: libc-alpha at sourceware dot org
- Date: Wed, 15 May 2013 11:47:26 -0400
- Subject: Re: [PING][PATCH][BZ #15022] Correct global-scope dlopen issues in static executables
- References: <alpine dot DEB dot 1 dot 10 dot 1301152056590 dot 4834 at tp dot orcam dot me dot uk> <20130116215545 dot 7A37A2C0B0 at topped-with-meat dot com> <alpine dot DEB dot 1 dot 10 dot 1301240655220 dot 4834 at tp dot orcam dot me dot uk> <alpine dot DEB dot 1 dot 10 dot 1305140020480 dot 26443 at tp dot orcam dot me dot uk> <5191C135 dot 9090601 at redhat dot com> <alpine dot DEB dot 1 dot 10 dot 1305151635591 dot 26443 at tp dot orcam dot me dot uk>
On 05/15/2013 11:37 AM, Maciej W. Rozycki wrote:
> On Tue, 14 May 2013, Carlos O'Donell wrote:
>
>> I've reviewed your code, and your new test cases and they look
>> good to me.
>>
>> Roland's comment about preserving user-visible semantics doesn't
>> seem to apply in this case. This case has never worked as far as
>> I can tell which means there are no directly user-visible semantics
>> to preserve. I would say your patch is a big step forwards to
>> preserving the user-visible semantics from the dynamic application
>> and ensuring that the static application mostly behaves in the
>> same way. It is completely acceptable IMO that the global scope
>> should be empty for a static executable and that dlsym() should
>> return appropriate errors.
>>
>> The additional tststatic5 and tststatic6 tests are exactly the tests
>> that I would have wanted; they are minimal and express the behaviour
>> that we expect from the API without involving anything overly
>> complicated.
>>
>> Thanks for the detailed problem description, and thorough analysis.
>> It's always a pleasure to review your patches.
>>
>> I'm OK with version 2 going into 2.18.
>
> Thanks for your review and words of appreciation. Does anyone else have
> anything to add?
Roland's recent comments have been that he is busy until next week.
I wouldn't expect him to respond until then.
Given that I feel we have addressed Roland's comments I would say
check this in on May 23nd of nobody objects. That gives another week
for others to comment or Roland to respond.
Cheers,
Carlos.