broadcom/compiler: padding fixes to QPU assembly dumps
When there are dst/src modifiers it is pretty common that instructions
take too much space and lead to alignment issues that make code a lot
harder to read, so align the MUL and SIG columns a bit wider to avoid
this:
Before:
0x380021828003faa8 fmax rf2, rf42.abs, rf40.abs; nop
0x3800f186c503f0f0 fcmp.pushc -, rf3, rf48; nop
0x380c038b85b83282 fmax rf11, rf10, rf2; mov.ifa rf14, rf46
0x3800219ab503f359 and rf26, rf13, rf25; nop
0x3820f186c503f2f0 fcmp.pushc -, rf11, rf48; nop ; thrsw
0x382c013fb5b8368e and rf63, rf26, rf14; mov.ifa rf4, rf46; thrsw
0x38002185b503ffc4 and rf5, rf63, rf4 ; nop
0x38002186b503f141 and rf6, rf5, rf1 ; nop
0x382031873503f186 vfpack tlb, rf6, rf6; nop ; thrsw
0x380031873503f18f vfpack tlb, rf6, rf15; nop
0x38003186bb03f000 nop ; nop
After:
0x380021828003faa8 fmax rf2, rf42.abs, rf40.abs ; nop
0x3800f186c503f0f0 fcmp.pushc -, rf3, rf48 ; nop
0x380c038b85b83282 fmax rf11, rf10, rf2 ; mov.ifa rf14, rf46
0x3800219ab503f359 and rf26, rf13, rf25 ; nop
0x3820f186c503f2f0 fcmp.pushc -, rf11, rf48 ; nop ; thrsw
0x382c013fb5b8368e and rf63, rf26, rf14 ; mov.ifa rf4, rf46 ; thrsw
0x38002185b503ffc4 and rf5, rf63, rf4 ; nop
0x38002186b503f141 and rf6, rf5, rf1 ; nop
0x382031873503f186 vfpack tlb, rf6, rf6 ; nop ; thrsw
0x380031873503f18f vfpack tlb, rf6, rf15 ; nop
0x38003186bb03f000 nop ; nop
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13545>
b42f4b8809