On 27.01.2011 23:20, Ian Romanick wrote:
> On 01/26/2011 04:30 AM, Christoph Bumiller wrote:
> > The current copy propagation code would propagate TEMP[0].x from (6)
> > into TEMP[1].x from (8) in the following, which is clearly wrong:
>
> > 6: MOV TEMP[1].x, TEMP[0].
> > 7: MOV TEMP[0]
On 27.01.2011 23:20, Christoph Bumiller wrote:
> On 27.01.2011 07:34, Eric Anholt wrote:
>> On Wed, 26 Jan 2011 13:30:04 +0100, Christoph Bumiller
>> wrote:
>>> The current copy propagation code would propagate TEMP[0].x from (6)
>>> into TEMP[1].x from (8) in the following, which is clearly wron
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 01/27/2011 02:20 PM, Christoph Bumiller wrote:
> glsl: add test for copy propagation bug
I converted the test to shader_runner and pushed it as
glsl-copy-propagation-loop-1.
Thanks for the test case.
-BEGIN PGP SIGNATURE-
Version: GnuPG v1
On 27.01.2011 07:34, Eric Anholt wrote:
> On Wed, 26 Jan 2011 13:30:04 +0100, Christoph Bumiller
> wrote:
>> The current copy propagation code would propagate TEMP[0].x from (6)
>> into TEMP[1].x from (8) in the following, which is clearly wrong:
>>
>> 6: MOV TEMP[1].x, TEMP[0].
>> 7:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 01/26/2011 04:30 AM, Christoph Bumiller wrote:
> The current copy propagation code would propagate TEMP[0].x from (6)
> into TEMP[1].x from (8) in the following, which is clearly wrong:
>
> 6: MOV TEMP[1].x, TEMP[0].
> 7: MOV TEMP[0].x,
On Wed, 26 Jan 2011 13:30:04 +0100, Christoph Bumiller
wrote:
> The current copy propagation code would propagate TEMP[0].x from (6)
> into TEMP[1].x from (8) in the following, which is clearly wrong:
>
> 6: MOV TEMP[1].x, TEMP[0].
> 7: MOV TEMP[0].x, TEMP[0].
> 8: MOV TEMP[0
The current copy propagation code would propagate TEMP[0].x from (6)
into TEMP[1].x from (8) in the following, which is clearly wrong:
6: MOV TEMP[1].x, TEMP[0].
7: MOV TEMP[0].x, TEMP[0].
8: MOV TEMP[0].y, TEMP[1].
9: ADD TEMP[1].x, TEMP[0]., IMM[0].
10: MOV
