Commit Graph

13 Commits

Author SHA1 Message Date
Craig Topper d35bcbbb5d [Sema][X86] Consider target attribute into the checks in validateOutputSize and validateInputSize.
The validateOutputSize and validateInputSize need to check whether
AVX or AVX512 are enabled. But this can be affected by the
target attribute so we need to factor that in.

This patch moves some of the code from CodeGen to create an
appropriate feature map that we can pass to the function.

Differential Revision: https://reviews.llvm.org/D68627
2019-12-23 11:23:30 -08:00
Reid Kleckner eff08f4097 Revert "[Sema][X86] Consider target attribute into the checks in validateOutputSize and validateInputSize."
This reverts commit e1578fd2b7.

It introduces a dependency on Attr.h which I am removing from
ASTContext.h.
2019-12-06 15:42:14 -08:00
Craig Topper e1578fd2b7 [Sema][X86] Consider target attribute into the checks in validateOutputSize and validateInputSize.
The validateOutputSize and validateInputSize need to check whether
AVX or AVX512 are enabled. But this can be affected by the
target attribute so we need to factor that in.

This patch copies some of the code from CodeGen to create an
appropriate feature map that we can pass to the function. Probably
need some refactoring here to share more code with Codegen. Is
there a good place to do that? Also need to support the cpu_specific
attribute as well.

Differential Revision: https://reviews.llvm.org/D68627
2019-12-06 15:30:59 -08:00
Akira Hatanaka e867e422e2 [X86, inlineasm] Do not allow using constraint 'x' for a variable larger than
128-bit unless the target CPU supports AVX.

rdar://problem/11846140

llvm-svn: 218082
2014-09-18 21:58:54 +00:00
Hans Wennborg 3c619a43d5 [X86, inline-asm] Allow 256-bit wide operands for the 'x' constraints
The 'x' constraint is for "any SSE register", and GCC seems to include the
256-bit ymm registers in that concept.

llvm-svn: 218073
2014-09-18 20:24:04 +00:00
Akira Hatanaka 974131ea88 [X86, inlineasm] Check that the output size is correct for the given constraint.
llvm-svn: 218064
2014-09-18 18:17:18 +00:00
Akira Hatanaka 31c6d3b71e [X86, inline-asm] Check that the input size is correct for constraints R, q, Q,
S, D, A, y, x, f, t, and u.

This is a follow-up patch for r167717.

rdar://problem/11846140
rdar://problem/17476970

llvm-svn: 217994
2014-09-17 23:35:14 +00:00
Bill Wendling c4fc3a2ba5 Emit an error message instead of crashing when dereferencing an incomplete pointer type.
If the ASM statement is dereferencing an incomplete pointer type, issue an error
instead of crashing.
<rdar://problem/12700799>

llvm-svn: 177915
2013-03-25 21:09:49 +00:00
Bill Wendling b3b4a37138 Use RequireCompleteType() instead of isIncompleteType().
isIncompleteType() returns true or false for template types depending on whether
the type is instantiated yet. In this context, that's arbitrary. The better way
to check for a complete type is RequireCompleteType().

Thanks to Eli Friedman for noticing this!

<rdar://problem/12700799>

llvm-svn: 177768
2013-03-22 21:33:46 +00:00
Bill Wendling 0cc8bc4d63 That's causing an error.
llvm-svn: 167763
2012-11-12 22:14:27 +00:00
Bill Wendling b1da2cb3fd Don't test for incomplete types.
llvm-svn: 167761
2012-11-12 22:01:56 +00:00
Bill Wendling 039136664d Update testcase to show that we don't emit an error for sizes <= 32-bits.
llvm-svn: 167748
2012-11-12 21:13:35 +00:00
Bill Wendling 887b485dbe Check that the input size is correct for the given constraint.
The 'a', 'c', and 'd' constraints on i386 mean a 32-bit register. We cannot
place a 64-bit value into the 32-bit register. Error out instead of causing the
compiler to spew general badness.
<rdar://problem/12415959>

llvm-svn: 167717
2012-11-12 06:42:51 +00:00