This is the mail archive of the binutils@sources.redhat.com mailing list for the binutils project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [PATCH] CFI directives for GAS


Zack Weinberg <zack@codesourcery.com> writes:

> Hans-Peter Nilsson <hp@bitrange.com> writes:
>
>> On Thu, 15 May 2003, Michal Ludvig wrote:
>>> For now the architectures that have these new directives enabled are
>>> i386 and amd64.
>>>
>>> Commented example on the usage can be found here:
>>> http://www.logix.cz/~mic/devel/gas-cfi/example.s.html
>>>
>>> How do you like this idea? Any comments? Can I apply the patch?
>>
>> You asked for comments, so here goes my usual knee-jerk
>> reaction: Cool, but how do we know this works and doesn't
>> trivially break with anyone's next set of patches?  Where's the
>> test-suite?  Even one i386 or amd64-specific case from your
>> example code would be terrific.  Consider that test-cases for
>> new features are mandatory in GCC, and for IMHO a good reason.

> Also, you didn't add any documentation.  GAS already has too many
> undocumented directives.

Michal documented this in his example and asked for comments whether
everybody with the extension per se.

Testsuites and documentation are missing and should be added before
the final submission.  But for such a new concept, let's discuss first
whether it's ok and serves everybody's need, looks implemented
correctly - and then do the rest.

I'm all in favor of adding such extension after having tried to add
eh_frame info by hand to glibc :-(

Andreas
-- 
 Andreas Jaeger
  SuSE Labs aj@suse.de
   private aj@arthur.inka.de
    http://www.suse.de/~aj


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]