[PATCH] 2nd round of Mips ifunc support

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

[PATCH] 2nd round of Mips ifunc support

Jack Carter
Attached is the second pass for Mips ifunc support in glibc.

There is an abi attached as well.

Jack

ifunc_abi.txt (11K) Download Attachment
mips_ifunc_glibc.patch (11K) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: [PATCH] 2nd round of Mips ifunc support

Will Newton
On 16 November 2013 00:41, Jack Carter <[hidden email]> wrote:
> Attached is the second pass for Mips ifunc support in glibc.
>
> There is an abi attached as well.

Your patch would be easier to review if it was posted inline with e.g.
git-send-email.

I don't think the vxworks comment in the commit message (or indeed the
reference to micromips etc.) makes any sense in the context of glibc.

The use of __builtin_expect should be replaced with __glibc_unlikely.

The example in your ABI doc should probably take an argument of the
hwcap value rather than void.

--
Will Newton
Toolchain Working Group, Linaro
Reply | Threaded
Open this post in threaded view
|

RE: [PATCH] 2nd round of Mips ifunc support

Jack Carter
Will, Thanks for looking at this. My responses are below.

Jack
________________________________________

> From: Will Newton [[hidden email]]
> Sent: Saturday, November 16, 2013 3:06 AM
> To: Jack Carter
> Cc: [hidden email]; [hidden email]
> Subject: Re: [PATCH] 2nd round of Mips ifunc support
>
> On 16 November 2013 00:41, Jack Carter <[hidden email]> wrote:
> > Attached is the second pass for Mips ifunc support in glibc.
> >
> > There is an abi attached as well.
>
> Your patch would be easier to review if it was posted inline with e.g.
> git-send-email.

Will do. I got reamed for using it for LLVM, but am learning the differences between
the 2 cultures.

>
> I don't think the vxworks comment in the commit message (or indeed the
> reference to micromips etc.) makes any sense in the context of glibc.

I will take them out.

>
> The use of __builtin_expect should be replaced with __glibc_unlikely.

I will make the changes.

>
> The example in your ABI doc should probably take an argument of the
> hwcap value rather than void.

I will make the changes.

>
> --
> Will Newton
> Toolchain Working Group, Linaro