Bring the -mcpu=power8 flag #7002
Conversation
get back the power8 flag to build two IBs
this was the reasoning behind the ppc64le:
```
-mtune=cpu_type
Set the instruction scheduling parameters for machine type cpu_type, but do not set the architecture type or register usage, as -mcpu=cpu_type does. The same values for cpu_type are used for -mtune as for -mcpu. If both are specified, the code generated uses the architecture and registers set by -mcpu, but the scheduling parameters set by -mtune.
```
the -mcpu=power8 specifies power8 machine which is more "generic" while the ppc64le is more specific for our machines.
If we read the following:
`The other options specify a specific processor. Code generated under those options runs best on that processor, and may not run at all on others. `
it seems to me the best flags to use will be those specifying the processor model (if it's in that list)
https://gcc.gnu.org/onlinedocs/gcc/RS_002f6000-and-PowerPC-Options.html#RS_002f6000-and-PowerPC-Options
|
A new Pull Request was created by @mrodozov (Mircho Rodozov) for branch IB/CMSSW_12_0_X/master. @cmsbuild, @smuzaffar, @mrodozov, @iarspider can you please review it and eventually sign? Thanks. |
|
test parameters:
|
|
please test |
cmsbuild
added
tests-started
tests-pending
and removed
tests-pending
tests-started
labels
|
abort |
|
please test |
|
now when I think about it ppc64le doesn't says which generation 7,8 or 9. maybe power8 is best (and all examples are using the -mcpu=power7,8,9 as argument) |
|
https://developer.ibm.com/technologies/linux/articles/compilers-for-lop-optimization-options/ |
|
please test |
|
please test for CMSSW_12_0_X/slc7_ppc64le_gcc9 |
|
please test for CMSSW_12_0_X/slc7_ppc64le_gcc9 |
|
-1 Failed Tests: UnitTests RelVals-THREADING The following merge commits were also included on top of IB + this PR after doing git cms-merge-topic:
You can see more details here: Unit TestsI found errors in the following unit tests: ---> test test_PrepareInputDb had ERRORS ---> test test_MpsWorkFlow had ERRORS ---> test DFM_MathRepr had ERRORS ---> test cudaMemUtils_t had ERRORS and more ... RelVals-THREADING
|
get back the power8 flag to build two IBs
this was the reasoning behind the ppc64le:
the -mcpu=power8 specifies power8 machine which is more "generic" while the ppc64le is more specific for our machines.
If we read the following:
The other options specify a specific processor. Code generated under those options runs best on that processor, and may not run at all on others.it seems to me the best flags to use will be those specifying the processor model (if it's in that list)
https://gcc.gnu.org/onlinedocs/gcc/RS_002f6000-and-PowerPC-Options.html#RS_002f6000-and-PowerPC-Options