Bug 90348

Summary: Spilling failure of b96 merged value
Product: Mesa Reporter: Ilia Mirkin <imirkin>
Component: Drivers/DRI/nouveauAssignee: Nouveau Project <nouveau>
Status: RESOLVED FIXED QA Contact: Nouveau Project <nouveau>
Severity: normal    
Priority: medium CC: r9ku1q
Version: git   
Hardware: Other   
OS: All   
Whiteboard:
i915 platform: i915 features:
Attachments: shader causing spill failure

Description Ilia Mirkin 2015-05-06 18:41:59 UTC 
Sometimes the spilling logic will decide to spill a merged def. This is fine for 2- and 4-wide values as there are appropriate load/store instructions for those, but not so with 3-wide values. We must take that into account in the SpillCodeInserter and emit a 64-bit and 32-bit store into lmem (and similar for load, although less important, as it tends to have gotten split up by the time the use rolls around).

Unfortunately I've lost track of the program that repro'd this issue.
Comment 1 Ilia Mirkin 2015-05-06 23:12:47 UTC 
Created attachment 115607 [details]
shader causing spill failure

Compile with -a e4 to get the failure. Works on GK110 since that has 256 registers.

This is what ultimately causes the failure at emit time:

140: texfetch 2D $r0 $s0 f32 $r48t $r48q (8)
141: texbar - # $r48t (8)
142: st b96 # l[0x0] $r48t (8)

codegen/nv50_ir_emit_nvc0.cpp:1659:emitLoadStoreType: Assertion `!"invalid type"' failed.
Comment 2 Ilia Mirkin 2015-05-26 17:27:52 UTC 
This is a preliminary patch to resolve the issue that I wrote a while back:

https://github.com/imirkin/mesa/commit/25c74a866b5b555699520552caeccb1e8426d86b.patch

Note that it needs a counterpart unspill variant too (although that's not hit by this shader). And ideally it'd be smart enough to go for a 64 + 32 store rather than 3x 32, but that's probably too much -- split then re-merge. (Or can split take unevenly sized defs? That seems like asking for trouble.)
Comment 4 Ilia Mirkin 2015-11-23 05:27:13 UTC 
I've pushed out a version of this patch. Should be included in 11.0.x and the upcoming 11.1 release.

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.