Bug 111737 - nir_opt_algebraic.c takes too long to compile
Summary: nir_opt_algebraic.c takes too long to compile
Status: RESOLVED MOVED
Alias: None
Product: Mesa
Classification: Unclassified
Component: glsl-compiler (show other bugs)
Version: git
Hardware: Other All
: not set not set
Assignee: Samuel Iglesias Gonsálvez
QA Contact: Intel 3D Bugs Mailing List
URL:
Whiteboard:
Keywords: bisected
Depends on:
Blocks:
 
Reported: 2019-09-18 15:24 UTC by Ian Romanick
Modified: 2019-09-20 13:47 UTC (History)
0 users

See Also:
i915 platform:
i915 features:


Attachments

Description Ian Romanick 2019-09-18 15:24:32 UTC
One our 72-thread build machine, release builds of Mesa used to take 90 seconds.  They now take closer to 8 minutes.  All of the extra time is spent compiling nir_opt_algebraic.c.  Debug builds do not seem to be affected.  I bisected this to:

f097247dd831da9b6e48baebc8b91efec3afcd28 is the first bad commit
commit f097247dd831da9b6e48baebc8b91efec3afcd28
Author: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Date:   Wed Dec 12 16:29:13 2018 +0100

    nir/algebraic: disable inexact optimizations depending on float controls execution mode
    
    If FLOAT_CONTROLS_SIGNED_ZERO_INF_NAN_PRESERVE or
    FLOAT_CONTROLS_DENORM_FLUSH_TO_ZERO are enabled, do not apply the
    inexact optimizations so the VK_KHR_shader_float_controls execution
    mode is respected.
    
    v2:
    - Do not apply inexact optimizations if SHADER_DENORM_FLUSH_TO_ZERO is
      enabled (Andres).
    
    v3:
    - Updated to renamed shader info member (Andres).
    
    v4:
    - Directly access execution mode instead of dragging it by parameter (Caio).
    
    Signed-off-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
    Signed-off-by: Andres Gomez <agomez@igalia.com>
    Reviewed-by: Connor Abbott <cwabbott0@gmail.com> [v1]

:040000 040000 33e076aacc5a6a4f10065cac2c8b06af0a5439a2 2b4a6e086b09648fc67a1f219777d51dc96f60f5 M	src
Comment 1 GitLab Migration User 2019-09-18 19:46:40 UTC
-- GitLab Migration Automatic Message --

This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity.

You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/mesa/mesa/issues/825.
Comment 2 Dieter Nützel 2019-09-20 13:47:22 UTC
I had the same issues.
~13 minutes, down to former ~4 minutes on my system  (4c/8t)

SOLVED with current master.

I do NOT have a GitLab account, currently.


Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.