1246619 - NSS clang-format: lib/freebl

Status: RESOLVED FIXED

(NSS :: Libraries, defect)

Description

[Franziskus Kiefer [:franziskus]]

Applying clang-format to lib/freebl.

Comment 1

[Franziskus Kiefer [:franziskus]]

Attachment: LibFreeblClangFormat.patch

clang-format on lib/freebl. patch is also at https://codereview.appspot.com/290100043

this is a big one. And for some reason clang-format did many strange line breaks. I tried to fix them but not sure if I spotted all. The patch also contains some clang-format protections. There was other code with special formatting, but I preferred the new format.

Comment 2

[Kai Engert (:KaiE:)]

It would be nice to make progress on this one.

Comment 3

[Franziskus Kiefer [:franziskus]]

Attachment: freebl-cf.patch

so this is a big one again.

There was quite some custom formatting in here (a lot of the algorithms), but nothing I found worthy of keeping.

Comment 4

[Franziskus Kiefer [:franziskus]]

Attachment: freebl-cf-ci.patch

this patch enables clang-format on freebl on TC CI

Comment 5

[Franziskus Kiefer [:franziskus]]

Comparison of freebl libs shows they're identical with and without the patch.

Comment 6

[Kai Engert (:KaiE:)]

Comment on attachment 8785241 [details] [diff] [review]
freebl-cf.patch

The only unexpected change I saw, clang-format replaces tabs with spaces inside of strings!

In two places, it's done for the usage string of the ecp_test/ecperf tools. I think we don't need to worry about the small change of the help/usage output this change will cause.

In one place it replaced the string given to an __asm("...") statement.
  __asm("bswap{tab}%0....);

I seems that doesn't change the resulting, so it's ok with me.

r=kaie

Comment 7

[Franziskus Kiefer [:franziskus]]

(In reply to Kai Engert (:kaie) from comment #6)
> Comment on attachment 8785241 [details] [diff] [review]
> freebl-cf.patch
> 
> The only unexpected change I saw, clang-format replaces tabs with spaces
> inside of strings!
> 
> In two places, it's done for the usage string of the ecp_test/ecperf tools.
> I think we don't need to worry about the small change of the help/usage
> output this change will cause.
> 
> In one place it replaced the string given to an __asm("...") statement.
>   __asm("bswap{tab}%0....);
> 
> I seems that doesn't change the resulting, so it's ok with me.
> 
> r=kaie

Thanks Kai! right, the changes in __asm don't change anything and there aren't actually any output strings that are used (ecp_test is dead code).

Comment 8

[Franziskus Kiefer [:franziskus]]

https://hg.mozilla.org/projects/nss/rev/d96445d2f8fd
https://hg.mozilla.org/projects/nss/rev/d36a5c822591