Skip to content

Sycl with unique stable name#3672

Closed
erichkeane wants to merge 43 commits intointel:syclfrom
erichkeane:SYCL_WITH_UNIQUE_STABLE_NAME
Closed

Sycl with unique stable name#3672
erichkeane wants to merge 43 commits intointel:syclfrom
erichkeane:SYCL_WITH_UNIQUE_STABLE_NAME

Conversation

@erichkeane
Copy link
Copy Markdown
Contributor

Just in place for validating.

@schittir and @AaronBallman , jsut FYI, no reason to review.

Erich Keane added 30 commits April 30, 2021 06:07
'Expr' type, which required a ton of implementation... Whew, at least
that is over.  A couple of 'TODO's remain, but we'll get to those
eventually.
… the name/mark the name, and some other various TODOs there
…e in both SEMA and CodeGen without duplication
…mbda' registration is a little wrong right now though
I was trying to use the same lambda to do 2 different things, determine
whether the lambda needed this mangling, and doing the mangling.
Unfortuantely the logic/split there was not particularly useful.

Changed back to have these function-pointers always be populated and
switched to 2 different calls.  This allows us to not have a conflict on
not-participate-in-kernel with lambda ID #0 and ID#1 (0 is supposed to
be mangled as an unnamed struct, and #1 is mangled with id 0).
This just makes it not particularly usable in the clang-c interface,
which I believe we're OK with.  In the future, we can re-add this
behavior if we care to.
The load was doing some funny stuff trying to turn the type of our
builtin into an i8 instead of an i8*.  However, it seems that when we
use a global we typically just want to wrap it in a GEP to get the 1st
element.  This is auto-generated for us if we jsut return the string
itself.
…rrect AS, then addr-space-cast if necessary
@romanovvlad
Copy link
Copy Markdown
Contributor

/summary:run

@romanovvlad
Copy link
Copy Markdown
Contributor

/summary:all

@erichkeane erichkeane closed this Jun 3, 2021
iclsrc pushed a commit that referenced this pull request Apr 10, 2026
A recent version of SPIRV-Tools found several issues with the test, such
as `DebugTypeFunction` having the wrong return type operand and
`DebugTypeBasic` missing the flags operand.

Original commit:
KhronosGroup/SPIRV-LLVM-Translator@bf469923a25d484
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants