Re: [RFC/PATCH v1 0/4] compat/obstack: update from upstream

From: Ramsay Jones <ramsay@ramsayjones.plus.com>
Date: Jun 14, 2019, 21:24 UTC
Message-ID: <b99ab7c9-63e3-6f2d-979b-d24a2670f9d5@ramsayjones.plus.com>
In-Reply-To: <bd955bb3-626e-345d-550e-562027c2f7e6@ramsayjones.plus.com>

On 14/06/2019 21:30, Ramsay Jones wrote:

Show 38 quoted lines

> 
> 
> On 14/06/2019 11:00, SZEDER Gábor wrote:
>> Update 'compat/obstack.{c,h}' from upstream, because they already use
>> 'size_t' instead of 'long' in places that might eventually end up as
>> an argument to malloc(), which might solve build errors with GCC 8 on
>> Windows.
>>
>> The first patch just imports from upstream and doesn't modify anything
>> at all, and, consequently, it can't be compiled because of a screenful
>> or two of errors.  This is bad for future bisects, of course.
>>
>> OTOH, adding all the necessary build fixes right away makes review
>> harder...
>>
>> I'm not sure how to deal with this situation, so here is a series with
>> the fixes in separate patches for review, for now.  If there's an
>> agreement that this is the direction to take, then I'll squash in the
>> fixes in the first patch and touch up the resulting commit message.
>>
>>
>> Ramsay, could you please run sparse on top of these patch series to
>> make sure that I caught and converted all "0 instead of NULL" usages
>> in the last patch?  Thanks.
> 
> I applied your patches to current master (@0aae918dd9) and, since
> you dropped the final hunk of commit 3254310863 ("obstack.c: Fix
> some sparse warnings", 2011-09-11), sparse complains, thus:
> 
>   $ diff sp-out sp-out1
>   27a28,30
>   > compat/obstack.c:331:5: warning: incorrect type in initializer (different modifiers)
>   > compat/obstack.c:331:5:    expected void ( *[addressable] [toplevel] obstack_alloc_failed_handler )( ... )
>   > compat/obstack.c:331:5:    got void ( [noreturn] * )( ... )
>   $ 
> 
> So, yes you did catch all "using plain integer as NULL pointer"
> warnings! :-D

Sorry for being a bit slow here, but I just realized that I should not have seen that on Linux (and should have tested on cygwin), because the obstack code gets elided on Linux ...

Oh, wait:

  $ diff sc sc1
  3a4,7
  > compat/obstack.o	- _obstack_allocated_p
  > compat/obstack.o	- obstack_alloc_failed_handler
  > compat/obstack.o	- _obstack_begin_1
  > compat/obstack.o	- _obstack_memory_used
  $

Hmm, so on master, this code is totally elided on Linux, but that is no longer the case with your patches applied. I guess you need to look at the definition of the {_OBSTACK_}ELIDE_CODE preprocessor variable(s).

HTH.

ATB, Ramsay Jones