This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.
Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
---|---|---|
Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |
Other format: | [Raw text] |
The enclosed patch fixes the two bugs noted in 41787, namely: 1) Vectors registers that overlap with the floating point registers cannot be saved across calls. 64-bit scalar values can be saved, but not the 128-bit vectors. This affected the spec 2006 benchmarks dealII, caculix, and GemsFDTD. 2) If the code used vectors that overlap with the floating point registers, but not any that overlap with the Altivec registers, the compiler would not set the VRSAVE register to non-zero, and it would be possible that the registers might not be saved by the kernel during context swaps. I have done the usual bootstrap and make check and there were no regressions. Is this ok to install? -- Michael Meissner, IBM 4 Technology Place Drive, MS 2203A, Westford, MA, 01886, USA meissner@linux.vnet.ibm.com
Attachment:
gcc-power7.patch86
Description: Text document
Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
---|---|---|
Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |