[Mesa-dev] [PATCH 2/2] amd/addrlib: update Mesa's copy of addrlib

[Bas Nieuwenhuizen]

So I tried to test this with radv and got a bunch of crashes in CTS,
mostly around 3d image support:

#3  0x00007ffff71a9396 in __assert_fail () from /usr/lib/libc.so.6
#4  0x00007ffff69da3b4 in
Addr::V2::Gfx9Lib::HwlGetPreferredSurfaceSetting (this=0x555557661b30,
pIn=0x7fffffffd5f0, pOut=0x7fffffffd5d0)
    at ../mesa/src/amd/addrlib/src/gfx9/gfx9addrlib.cpp:3684
#5  0x00007ffff69cf331 in
Addr::V2::Lib::Addr2GetPreferredSurfaceSetting (this=0x555557661b30,
pIn=0x7fffffffd5f0, pOut=0x7fffffffd5d0)
    at ../mesa/src/amd/addrlib/src/core/addrlib2.cpp:1742
#6  0x00007ffff69c4e87 in Addr2GetPreferredSurfaceSetting
(hLib=0x555557661b30, pIn=0x7fffffffd5f0, pOut=0x7fffffffd5d0)
    at ../mesa/src/amd/addrlib/src/addrinterface.cpp:1697
#7  0x00007ffff69bf8d4 in gfx9_get_preferred_swizzle_mode
(addrlib=0x555557661b30, in=0x7fffffffd690, is_fmask=false,
flags=33555202, swizzle_mode=0x7fffffffd698)

It seems to be caused by the explicit swizzle mode override that we do with

commit b64b7125586ce48232658cd860f549a6139b6ddd
Author: Marek Olšák <marek.olsak at amd.com>
Date:   Mon Apr 2 12:54:52 2018 -0400

    ac/surface/gfx9: request desired micro tile mode explicitly

    Tested-by: Dieter Nützel <Dieter at nuetzel-hh.de>


Since we never got a reason to have it (the commit message above is
not descriptive and the patch not reviewed) and this is the second
time already that this breaks stuff (The other was allowing S tiling
for raven displayable surfaces, per 7eff8d7d3564), maybe revert it and
let addrlib make the decision?