- User Since
- Jan 26 2017, 7:07 AM (191 w, 3 d)
Jul 8 2020
May 27 2020
I am abandoning this, because it turns out that it is difficult to reliably detect whether we have a PAL pipeline or a PAL shader compilation. Instead, the LLPC linker will remove the padding, and then adds it back on after concatenating .text sections.
May 26 2020
May 25 2020
V2: Review comments: Separate predicate function; early exit on CS.
May 21 2020
V5: API change so client can choose whether array merging does merge or append.
May 20 2020
V4: Added MapDocNode::erase methods.
V2: Fixed tests.
May 19 2020
May 18 2020
V3: Added assignment for const char *, otherwise it defaults to using
the bool one rather than the StringRef one.
V2: Added tests
With the nil/empty change, there is a change in behavior in that doing a map lookup "map[key]" with a key not previously in the map creates a new map entry that is empty rather than nil. That is more correct, but I have just discovered some code in LLPC relying on the old behavior. I can fix the LLPC code, but is the change in behavior going to cause any HSA problems?
V4: Addressed review comments: * Separated the idea of an empty node from a nil node * Made default merge behavior to fail on any conflict * Array merging now appends
May 13 2020
May 12 2020
May 11 2020
V3: Removed unneeded nullptr arg.
May 10 2020
May 9 2020
V2: Pass map key to merger function.
May 1 2020
V2: Also fixed GS regs. Added test.
Apr 25 2020
Mar 10 2020
I don't know of any reason why it has to be fp. Maybe related to the prehistoric thing where isel used to need something to be fp to put it in a vgpr?
Feb 17 2020
This got landed in a slightly different form because of conflicts with other sdiv64 etc tests.
Feb 14 2020
V2: Also added gfx1010 to memtime test.
Dec 13 2019
V3: Reinstated and fixed the removed test.
Dec 12 2019
Yes, sorry, I will do the test changes that Matt suggested so this can be approved and landed.
Dec 4 2019
V2: Removed an existing test for the bogus fold.
Dec 3 2019
Dec 2 2019
Nov 27 2019
V2: Sorted includes alphabetically.
No test because by definition it was indeterminate.
Nov 12 2019
Nov 11 2019
V3: Only bail if it is a copy and a partial def, to avoid spurious test changes.
Nov 8 2019
V2: Used modifiesRegister as suggested by Matt.
This is slightly broken for us. See D70005 for a fix.
Nov 7 2019
Oct 29 2019
Sep 30 2019
V2: Better fix that does not accidentally allow pow() transforms.
@foad pointed out that this fix is wrong. TLI saying pow() is not supported means if we find a call to a function called pow() then we don't know its semantics. So I will push a revised fix.
Sep 23 2019
Sep 18 2019
Sep 17 2019
Sep 11 2019
Sep 10 2019
V2: Disable all library functions, not just tan.