[Mesa-dev] [PATCH 06/10] i965/fs: Fix dump_instructions() on uniforms.
Eric Anholt
eric at anholt.net
Thu Mar 27 16:12:57 PDT 2014
Kenneth Graunke <kenneth at whitecape.org> writes:
> On 03/26/2014 02:23 PM, Eric Anholt wrote:
>> All of a vec4 uniform was being printed as "u0"
>> ---
>> src/mesa/drivers/dri/i965/brw_fs.cpp | 4 ++--
>> 1 file changed, 2 insertions(+), 2 deletions(-)
>>
>> diff --git a/src/mesa/drivers/dri/i965/brw_fs.cpp b/src/mesa/drivers/dri/i965/brw_fs.cpp
>> index 713e477..b3f1dfd 100644
>> --- a/src/mesa/drivers/dri/i965/brw_fs.cpp
>> +++ b/src/mesa/drivers/dri/i965/brw_fs.cpp
>> @@ -3035,7 +3035,7 @@ fs_visitor::dump_instruction(backend_instruction *be_inst)
>> fprintf(stderr, "(null)");
>> break;
>> case UNIFORM:
>> - fprintf(stderr, "***u%d***", inst->dst.reg);
>> + fprintf(stderr, "***u%d***", inst->dst.reg + inst->dst.reg_offset);
>> break;
>> case HW_REG:
>> if (inst->dst.fixed_hw_reg.file == BRW_ARCHITECTURE_REGISTER_FILE) {
>> @@ -3087,7 +3087,7 @@ fs_visitor::dump_instruction(backend_instruction *be_inst)
>> fprintf(stderr, "***m%d***", inst->src[i].reg);
>> break;
>> case UNIFORM:
>> - fprintf(stderr, "u%d", inst->src[i].reg);
>> + fprintf(stderr, "u%d", inst->src[i].reg + inst->src[i].reg_offset);
>> if (inst->src[i].reladdr) {
>> fprintf(stderr, "+reladdr");
>> } else if (virtual_grf_sizes[inst->src[i].reg] != 1 ||
>
> This kind of makes me wonder whether we should be using the reg_offset
> field for UNIFORM files at all. It seems like we universally add them
> together, and there's a bunch of optimization code that bails when
> reg_offset != 0...
Yeah, I've been thinking about switching uniforms to not use reg_offset,
as a step toward eliminating reg_offset completely. But until we get
around to that, I figured at least fixing the debug would be nice.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 818 bytes
Desc: not available
URL: <http://lists.freedesktop.org/archives/mesa-dev/attachments/20140327/6971a8fe/attachment.sig>
More information about the mesa-dev
mailing list