On Sat, Aug 6, 2016 at 4:34 AM, Rob Clark <robdcl...@gmail.com> wrote:
> On Fri, Aug 5, 2016 at 8:42 PM, Jan Ziak <0xe2.0x9a.0...@gmail.com> wrote:
>> Mesa source code prior to this patch uses both RTLD_NOW and RTLD_LAZY.
>> This patch removes all RTLD_NOW in favor of RTLD_LAZY.
>>
>> In comparison to early binding, lazy binding reduces CPU instruction count
>> of small GL apps (e.g: glxinfo) by 6 million instructions.
>> Larger apps won't notice the difference.
>
> tbh, I don't know the background of existing places that use RTLD_LAZY
> instead of RTLD_NOW (but my experience w/ xserver using LAZY has not
> been positive, so I think going the other direction seems like a good
> idea)..
We could add a verifier to the build process that tests the foo_dri.so
libraries (as well as all other libs subject to dlopen by Mesa) for
undefined symbols:
$ LD_PRELOAD=libGL.so ldd -d -r radeonsi_dri.so \
| grep "^undefined" && echo "red alert!"
This will ensure that Mesa does not break apps after replacing all
RTLD_NOWs with RTLD_LAZY.
I am going to start writing a new patch verifying relevant *.so files
at buildtime.
> But I'm not sure that optimizing for glxinfo is the best goal.
I will turn my attention to manifestations of suboptimal Mesa GL in
other apps as soon as I am satisfied with glxinfo/glxgears (and after
the list of my patches at patchwork gets closer to zero).
_______________________________________________
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev
