The search and replace function in the editor becomes unstable after several hours of work on a big file (> 2000 lines). The "replace all" feature replaces only the first occurrence of the search text…
Created on 4/7/2012 (updated 32 weeks ago) | 0 validations | 0 workarounds | 3 comments | feedback id: 735862
|
|
|
Closed
as Fixed
|
|
The attached project gives a fatal error C1001 when compiling for win32, release mode.
The error does not occur when compiling for debug mode or x64.
The project includes XOP and AVX2 intrinsics.
VS…
Created on 4/7/2012 (updated 28 weeks ago) | 0 validations | 0 workarounds | 5 comments | feedback id: 735861
|
|
|
Closed
as Fixed
|
|
VS 11 beta, C++ compiler Version 17.00.50214.1
The AVX2 intrinsics _mm256_permutevar8x32_epi32 and _mm256_permutevar8x32_ps defined in immintrin.h are putting the operands in wrong order when comp…
Created on 4/4/2012 (updated 58 weeks ago) | 0 validations | 0 workarounds | 3 comments | feedback id: 735397
|
|
|
Closed
as Duplicate
|
|
C++ Incorrect optimization after _mm256_castps128_ps256 and similar casts. Only 128 bits of the 256-bit register are stored.
Created on 3/20/2012 (updated 59 weeks ago) | 0 validations | 0 workarounds | 3 comments | feedback id: 732119
|
|
|
Closed
as By Design
|
|
During debugging of code containing AVX instructions, xmm registers are erroneously shown as ymm registers.
Created on 11/8/2009 (updated 183 weeks ago) | 0 validations | 0 workarounds | 3 comments | feedback id: 509069
|
|
|
Closed
as Fixed
|
|
The assembler ML.EXE and ML64.EXE appears to support the AVX instruction set, but the 32 bit version is missing a way to enable ymm registers:
Example:
ML.EXE
---------
.686
.xmm
.model flat…
Created on 11/8/2009 (updated 63 weeks ago) | 0 validations | 0 workarounds | 3 comments | feedback id: 509067
|
|
|
Closed
as Not Reproducible
|
|