intel/llvm
1
0
Fork 0
mirror of https://github.com/intel/llvm.git synced 2026-01-18 07:57:36 +08:00
Code Issues Packages Projects Releases Wiki Activity

Do not assert on shifts of Neon polynomial types.

Browse Source 
Most Neon shift intrinsics do not have variants for polynomial types, but
vsri_n and vsli_n do support them, and we need to properly range-check the
shift immediates for them.

llvm-svn: 121509
This commit is contained in:
Bob Wilson
2010-12-10 19:45:06 +00:00
parent fb0a3754f5
commit a880fa0b28
1 changed files with 2 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
clang/lib/Sema/SemaChecking.cpp
View File

@@ -198,11 +198,9 @@ static unsigned RFT(unsigned t, bool shift = false) {
assert(!shift && "cannot shift float types!");
return (2 << (int)quad) - 1;
case 5: // poly8
assert(!shift && "cannot shift polynomial types!");
return (8 << (int)quad) - 1;
return shift ? 7 : (8 << (int)quad) - 1;
case 6: // poly16
assert(!shift && "cannot shift polynomial types!");
return (4 << (int)quad) - 1;
return shift ? 15 : (4 << (int)quad) - 1;
case 7: // float16
assert(!shift && "cannot shift float types!");
return (4 << (int)quad) - 1;