efriedma added a comment.
There's nothing wrong with this change, as far as I can tell. That said, if
you're planning to use getRealTypeByWidth anywhere other than AddModeAttr, it's
probably a bug; the behavior of getRealTypeByWidth isn't useful for any other
purpose given that clang supports
ddcc updated this revision to Diff 79996.
ddcc added a comment.
Change definition
https://reviews.llvm.org/D26955
Files:
lib/Basic/TargetInfo.cpp
lib/Sema/SemaDeclAttr.cpp
Index: lib/Sema/SemaDeclAttr.cpp
===
--- lib/Sema/Sem
efriedma added inline comments.
Comment at: lib/Basic/TargetInfo.cpp:229
switch (BitWidth) {
- case 96:
+ case 80:
if (&getLongDoubleFormat() == &llvm::APFloat::x87DoubleExtended)
ddcc wrote:
> bruno wrote:
> > The change makes sense but I believe there
ddcc added inline comments.
Comment at: lib/Basic/TargetInfo.cpp:229
switch (BitWidth) {
- case 96:
+ case 80:
if (&getLongDoubleFormat() == &llvm::APFloat::x87DoubleExtended)
bruno wrote:
> The change makes sense but I believe there's some historical r
bruno added inline comments.
Comment at: lib/Basic/TargetInfo.cpp:229
switch (BitWidth) {
- case 96:
+ case 80:
if (&getLongDoubleFormat() == &llvm::APFloat::x87DoubleExtended)
The change makes sense but I believe there's some historical reason why this
ddcc added a comment.
I could be completely mistaken here, but currently
`Ctx.getRealTypeForBitwidth(llvm::APFloat::getSizeInBits(llvm::APFloat::x87DoubleExtended))`
with a `ASTContext Ctx` does not round-trip correctly.
https://reviews.llvm.org/D26955
__
ddcc created this revision.
ddcc added a reviewer: rsmith.
ddcc added a subscriber: cfe-commits.
llvm::APFloat::x87DoubleExtended is defined as having 80 bits of size
https://reviews.llvm.org/D26955
Files:
lib/Basic/TargetInfo.cpp
Index: lib/Basic/TargetInfo.cpp