Mesa (main): intel/compiler: Fix uncompaction of signed word immediates on Tigerlake

[GitLab Mirror]

Module: Mesa
Branch: main
Commit: 26bb81f3f62ff5a19f360f4b6d2f72bdcb229360
URL:    http://cgit.freedesktop.org/mesa/mesa/commit/?id=26bb81f3f62ff5a19f360f4b6d2f72bdcb229360

Author: Kenneth Graunke <kenneth at whitecape.org>
Date:   Thu Jun  2 02:11:15 2022 -0700

intel/compiler: Fix uncompaction of signed word immediates on Tigerlake

This expression accidentally performs a 32-bit sign-extension when
processing the second half of the expression (the low 16 bits).

Consider -7W, which is represented as 0xfff9fff9 in our encoding (the
16-bit word is replicated to both halves of the 32-bit dword).

Tigerlake's compaction stores the low 11-bits of an immediate as-is,
and replicates the 12th bit.  So here, compacted_imm will be 0xff9.

   (  (int)(0xff9 << 20) >> 4) |
   ((short)(0xff9 <<  4) >> 4))

   0xfff90000 | (0xff90 >> 4)
   0xfff90000 | 0xfffffff9 ...oops...
   0xfffffff9

By casting the second line of the expression to unsigned short, we
prevent the sign-extension when it combines both parts, so we get:

   0xfff90000 | 0x0000fff9
   0xfff9fff9

Fixes: 12d3b11908e ("intel/compiler: Add instruction compaction support on Gen12")
Reviewed-by: Matt Turner <mattst88 at gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/16833>

---

 src/intel/compiler/brw_eu_compact.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/src/intel/compiler/brw_eu_compact.c b/src/intel/compiler/brw_eu_compact.c
index 73fdbf5f8e0..376eddd99da 100644
--- a/src/intel/compiler/brw_eu_compact.c
+++ b/src/intel/compiler/brw_eu_compact.c
@@ -1658,8 +1658,8 @@ uncompact_immediate(const struct intel_device_info *devinfo,
          return (int)(compact_imm << 20) >> 20;
       case BRW_REGISTER_TYPE_W:
          /* Extend the 12th bit into the high 4 bits and replicate */
-         return (  (int)(compact_imm << 20) >> 4) |
-                ((short)(compact_imm <<  4) >> 4);
+         return ((int)(compact_imm << 20) >> 4) |
+                ((unsigned short)((short)(compact_imm << 4) >> 4));
       case BRW_REGISTER_TYPE_NF:
       case BRW_REGISTER_TYPE_DF:
       case BRW_REGISTER_TYPE_Q: