kevin/linux
1
Fork 0
Code

drm/amdkfd: Use S_ENDPGM_SAVED in trap handler

Browse Source 
This instruction has no functional difference to S_ENDPGM
but allows performance counters to track save events correctly.

Signed-off-by: Jay Cornwall <jay.cornwall@amd.com>
Reviewed-by: Laurent Morichetti <laurent.morichetti@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This commit is contained in:
Jay Cornwall 2024-01-05 08:49:06 -06:00 committed by Alex Deucher
parent 961df30854
commit c49bf4fcfc
3 changed files with 9 additions and 9 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h
View File

@ -674,7 +674,7 @@ static const uint32_t cwsr_trap_gfx9_hex[] = {
0x86ea6a6a, 0x8f6e837a, 0x86ea6a6a, 0x8f6e837a,
0xb96ee0c2, 0xbf800002, 0xb96ee0c2, 0xbf800002,
0xb97a0002, 0xbf8a0000, 0xb97a0002, 0xbf8a0000,
0xbe801f6c, 0xbf810000, 0xbe801f6c, 0xbf9b0000,
}; };
static const uint32_t cwsr_trap_nv1x_hex[] = { static const uint32_t cwsr_trap_nv1x_hex[] = {
@ -1091,7 +1091,7 @@ static const uint32_t cwsr_trap_nv1x_hex[] = {
0xb9eef807, 0x876dff6d, 0xb9eef807, 0x876dff6d,
0x0000ffff, 0x87fe7e7e, 0x0000ffff, 0x87fe7e7e,
0x87ea6a6a, 0xb9faf802, 0x87ea6a6a, 0xb9faf802,
0xbe80226c, 0xbf810000, 0xbe80226c, 0xbf9b0000,
0xbf9f0000, 0xbf9f0000, 0xbf9f0000, 0xbf9f0000,
0xbf9f0000, 0xbf9f0000, 0xbf9f0000, 0xbf9f0000,
0xbf9f0000, 0x00000000, 0xbf9f0000, 0x00000000,
@ -1574,7 +1574,7 @@ static const uint32_t cwsr_trap_arcturus_hex[] = {
0x86ea6a6a, 0x8f6e837a, 0x86ea6a6a, 0x8f6e837a,
0xb96ee0c2, 0xbf800002, 0xb96ee0c2, 0xbf800002,
0xb97a0002, 0xbf8a0000, 0xb97a0002, 0xbf8a0000,
0xbe801f6c, 0xbf810000, 0xbe801f6c, 0xbf9b0000,
}; };
static const uint32_t cwsr_trap_aldebaran_hex[] = { static const uint32_t cwsr_trap_aldebaran_hex[] = {
@ -2065,7 +2065,7 @@ static const uint32_t cwsr_trap_aldebaran_hex[] = {
0x86ea6a6a, 0x8f6e837a, 0x86ea6a6a, 0x8f6e837a,
0xb96ee0c2, 0xbf800002, 0xb96ee0c2, 0xbf800002,
0xb97a0002, 0xbf8a0000, 0xb97a0002, 0xbf8a0000,
0xbe801f6c, 0xbf810000, 0xbe801f6c, 0xbf9b0000,
}; };
static const uint32_t cwsr_trap_gfx10_hex[] = { static const uint32_t cwsr_trap_gfx10_hex[] = {
@ -2500,7 +2500,7 @@ static const uint32_t cwsr_trap_gfx10_hex[] = {
0x876dff6d, 0x0000ffff, 0x876dff6d, 0x0000ffff,
0x87fe7e7e, 0x87ea6a6a, 0x87fe7e7e, 0x87ea6a6a,
0xb9faf802, 0xbe80226c, 0xb9faf802, 0xbe80226c,
0xbf810000, 0xbf9f0000, 0xbf9b0000, 0xbf9f0000,
0xbf9f0000, 0xbf9f0000, 0xbf9f0000, 0xbf9f0000,
0xbf9f0000, 0xbf9f0000, 0xbf9f0000, 0xbf9f0000,
}; };
@ -2944,7 +2944,7 @@ static const uint32_t cwsr_trap_gfx11_hex[] = {
0xb8eef802, 0xbf0d866e, 0xb8eef802, 0xbf0d866e,
0xbfa20002, 0xb97af802, 0xbfa20002, 0xb97af802,
0xbe80486c, 0xb97af802, 0xbe80486c, 0xb97af802,
0xbe804a6c, 0xbfb00000, 0xbe804a6c, 0xbfb10000,
0xbf9f0000, 0xbf9f0000, 0xbf9f0000, 0xbf9f0000,
0xbf9f0000, 0xbf9f0000, 0xbf9f0000, 0xbf9f0000,
0xbf9f0000, 0x00000000, 0xbf9f0000, 0x00000000,
@ -3436,5 +3436,5 @@ static const uint32_t cwsr_trap_gfx9_4_3_hex[] = {
0x86ea6a6a, 0x8f6e837a, 0x86ea6a6a, 0x8f6e837a,
0xb96ee0c2, 0xbf800002, 0xb96ee0c2, 0xbf800002,
0xb97a0002, 0xbf8a0000, 0xb97a0002, 0xbf8a0000,
0xbe801f6c, 0xbf810000, 0xbe801f6c, 0xbf9b0000,
}; };

2
drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx10.asm
View File

@ -1104,7 +1104,7 @@ L_RETURN_WITHOUT_PRIV:
s_rfe_b64 s_restore_pc_lo //Return to the main shader program and resume execution s_rfe_b64 s_restore_pc_lo //Return to the main shader program and resume execution
L_END_PGM: L_END_PGM:
s_endpgm s_endpgm_saved
end end
function write_hwreg_to_mem(s, s_rsrc, s_mem_offset) function write_hwreg_to_mem(s, s_rsrc, s_mem_offset)

2
drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm
View File

@ -921,7 +921,7 @@ L_RESTORE:
/* the END */ /* the END */
/**************************************************************************/ /**************************************************************************/
L_END_PGM: L_END_PGM:
s_endpgm s_endpgm_saved
end end