git/list[1] front-page[2] threads[3] people[4] search[5] about
 

Re: [PATCH] hash: Allow building with the external sha1dc library

From
Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Date
Aug 1, 2017, 19:55 UTC
Message-ID
<87ini7m5a6.fsf@gmail.com>
In-Reply-To
<s5hmv7jw9lx.wl-tiwai@suse.de>
On Tue, Aug 01 2017, Takashi Iwai jotted:
Show 45 quoted lines
> On Tue, 01 Aug 2017 17:56:00 +0200,
> Junio C Hamano wrote:
>>
>> Takashi Iwai <tiwai@suse.de> writes:
>>
>> > On Fri, 28 Jul 2017 17:58:14 +0200,
>> > Ævar Arnfjörð Bjarmason wrote:
>> >>  ...
>> >> * We now have much of the same header code copy/pasted between
>> >> sha1dc_git.h and sha1dc_git_ext.h, did you consider just always
>> >> including the former but making what it's doing conditional on
>> >> DC_SHA1_EXTERNAL? I don't know if it would be worth it from a cursory
>> >> glance, but again your commit message doesn't list that among options
>> >> considered & discarded.
>> >
>> > I don't mind either way, there is no perfect solution in this case.
>> > As you know, many people think the ifdef ugly no matter how.
>> >
>> > I leave the decision to maintainer.  Just let me know which option is
>> > preferred.
>>
>> Yeah, I also found it somewhat confusing to have these two headers
>> that look quite similar to each other at the top-level of the tree.
>>
>> What's the "conditional" part between the two headers?  Is it just
>> whether the header for underlying library is included?  I wonder if
>> it's just the matter of adjusting "hash.h" to read like this
>>
>>     ...
>>     #if defined(DC_SHA1_EXTERNAL)
>>    -#include "sha1dc_git_ext.h"
>>    +#include <sha1dc/sha1.h>
>>    +#include "sha1dc_git.h"
>>     #elif  defined(DC_SHA1_SUBMODULE)
>>     ...
>>
>> or are there heavier tweaks needed that won't be solved by continuing
>> along the same line?  As _ext.h variant is included only at this place,
>> if we can do with minimum tweaks around here without introducing it,
>> it may be ideal, I would think.
>
> Well, a tricky part is that currently sha1dc_git.h is included from
> sha1dc/sha1.h implicitly by SHA1DC_CUSTOM_TRAILING_INCLUDE_SHA1_H
> definition.  IMO, we should stop this, and use the standard inclusion
> instead, i.e. in hash.h,

It's just like this because when I hacked up that facility I was making the bare minimum change needed to not make local modifications to the upstream code. If there's better ways to do this in the presence of DC_SHA1_EXTERNAL & without that would be most welcome. Thanks.

Show 14 quoted lines
> #if defined(DC_SHA1_EXTERNAL)
> #include <sha1dc/sha1.h>
> #elif defined(DC_SHA1_SUBMODULE)
> #include "sha1collisiondetection/lib/sha1.h"
> #else
> #include "sha1dc/sha1.h"
> #endif
> #include "sha1dc_git.h"
>
> In sha1dc_git.h, we'd need another ifdef for DC_SHA1_EXTERNAL to
> define the own git_SHA1DCInit(), but it's trivial.
>
>
> Takashi
Previous: Takashi IwaiNext: Takashi Iwai
Message 12 of 13 in “hash: Allow building with the external sha1dc library”
  1. hash: Allow building with the external sha1dc libraryTakashi Iwai, Jul 25, 2017
  2. Junio C HamanoJul 25, 2017
  3. Ævar Arnfjörð BjarmasonJul 28, 2017
  4. Ævar Arnfjörð BjarmasonJul 28, 2017
  5. Junio C HamanoJul 31, 2017
  6. Takashi IwaiAug 1, 2017
  7. Junio C HamanoAug 12, 2017
  8. Takashi IwaiAug 12, 2017
  9. Takashi IwaiAug 1, 2017
  10. Junio C HamanoAug 1, 2017
  11. Takashi IwaiAug 1, 2017
  12. Ævar Arnfjörð BjarmasonAug 1, 2017
  13. Takashi IwaiAug 1, 2017

Read the whole thread, see it on lore, or plain text.

$ cat FOOTERMessages come from the public archive at lore.kernel.org/git, fetched every hour. The front page is chosen and written each morning by an AI editor and can be wrong; the threads themselves are the record. About and API. For agents: an MCP server at https://gitlist.dev/mcp, and any thread, story or person page as Markdown by adding .md to its URL (or sending Accept: text/markdown). Details in /llms.txt.