https://llvm.org/bugs/show_bug.cgi?id=28165
Bug ID: 28165
Summary: [mc] F64 insns -- floating inline constants
incorrectly encoded as literals
Product: libraries
Version: trunk
Hardware: PC
OS: Windows NT
Status: NEW
Severity: normal
Priority: P
Component: Backend: AMDGPU
Assignee: artem.tama...@amd.com
Reporter: artem.tama...@amd.com
CC: llvm-bugs@lists.llvm.org, nikolay.haus...@amd.com
Classification: Unclassified
Example of failing instruction:
> v_ceil_f64 v[27:28], 0.5
In the case above, 0.5 is encoded as 32-bit literal (0x3f000000), which is
incorrect and will be interpreted as 3.0517578125E-5 instead of 0.5. Correct
encoding is inline constant 240 (optimal) or 32-bit literal 0x3FE00000 at
least.
I am almost sure that issue appears for other sub-targets as well, e.g. gfx8.
Number of failing lit1 tests (-search=simple) = 29349.
Tested with r272673 (git # da10a460d7c98278561225fda90855b303675f35).
--
You are receiving this mail because:
You are on the CC list for the bug.
_______________________________________________
llvm-bugs mailing list
llvm-bugs@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
