Skip to content

New ARM armv7a kernels and .S file support in Makefile#3

Merged
fgvanzee merged 2 commits intoflame:masterfrom
figual:master
Feb 21, 2014
Merged

New ARM armv7a kernels and .S file support in Makefile#3
fgvanzee merged 2 commits intoflame:masterfrom
figual:master

Conversation

@figual
Copy link
Copy Markdown
Collaborator

@figual figual commented Feb 21, 2014

Field,

(Just trying out the Pull Request functionality).

This request has two main modifications:

  1. Added a new config folder with the configuration files and micro-kernels from Werner Saar, named armv7a. We still need to discuss if this should replace the old cortex-a9 and cortex-a15 configs, as I am not seeing yet evindent improvements in S/D kernels. These configuration includes, though, tuned kernels for C/Z.
  2. Modified main Makefile in order to consider .S assembly files in the kernels folder. Note I'm not proposing this to replace the old Makefile, but to make you consider an elegant solution (better than mine) to support adding assembly micro-kernels in the framework.

Fran

fgvanzee added a commit that referenced this pull request Feb 21, 2014
New ARM armv7a kernels and Assembly file consideration in Makefile
@fgvanzee fgvanzee merged commit fc04b5e into flame:master Feb 21, 2014
@tlrmchlsmth
Copy link
Copy Markdown
Member

So who do these emails go to anyway? All the people who ‘follow’ the BLIS repo?

On Feb 21, 2014, at 7:04 AM, Field G. Van Zee [email protected] wrote:

Merged #3.


Reply to this email directly or view it on GitHub.

@songmaotian songmaotian mentioned this pull request Apr 22, 2016
@loveshack loveshack mentioned this pull request Mar 5, 2018
loveshack pushed a commit to loveshack/blis that referenced this pull request Sep 24, 2019
This needs fixing properly somehow, but using -O3 (at least with gcc 8.3),
we get this:

Program received signal SIGILL, Illegal instruction.
0x000000001004c660 in bli_cntx_init_power9_ref (cntx=0x103e06b0)
    at ref_kernels/bli_cntx_ref.c:456
456             for ( i = 0; i < BLIS_NUM_LEVEL3_OPS; ++i ) vfuncs[ i ] = NULL;
(gdb) bt
#0  0x000000001004c660 in bli_cntx_init_power9_ref (cntx=0x103e06b0)
    at ref_kernels/bli_cntx_ref.c:456
flame#1  0x000000001004c0a8 in bli_cntx_init_power9 (cntx=<optimized out>)
    at config/power9/bli_cntx_init_power9.c:42
flame#2  0x000000001003c85c in bli_gks_register_cntx (id=BLIS_ARCH_POWER9,
    nat_fp=0x1004c090 <bli_cntx_init_power9>,
    ref_fp=0x1004c0d0 <bli_cntx_init_power9_ref>, ind_fp=<optimized out>)
    at frame/base/bli_gks.c:373
flame#3  0x000000001003c97c in bli_gks_init () at frame/base/bli_gks.c:155
flame#4  0x000000001003cfe8 in bli_init_apis () at frame/base/bli_init.c:78
flame#5  0x00007ffff7e045a8 in __pthread_once_slow () from /lib64/libpthread.so.0
flame#6  0x00000000100492e8 in bli_pthread_once (once=<optimized out>,
    init=<optimized out>) at frame/thread/bli_pthread.c:314
flame#7  0x000000001003d138 in bli_init_once () at frame/base/bli_init.c:104
flame#8  bli_init_auto () at frame/base/bli_init.c:54
flame#9  0x0000000010011300 in cdotc_ (n=<optimized out>, x=<optimized out>,
    incx=<optimized out>, y=<optimized out>, incy=<optimized out>)
    at frame/compat/bla_dot.c:89
flame#10 0x0000000010002a48 in check2_ (sfac=0x103d14dc <sfac>)
    at blastest/src/cblat1.c:529
flame#11 0x0000000010001ef4 in main () at blastest/src/cblat1.c:112
niyas-sait pushed a commit to niyas-sait/blis that referenced this pull request Feb 25, 2022
* Update to new blis

* Remove gcc requirement from travis

* Update make log for Linux

* Fix make log

* Fix header

* Update make log

* Update linux build

* Update munge_make_log script

* Fix make log for gcc 4.8

* Add darwin build logs

* Use clang for darwin

* Add Windows build log

* Fix Windows make log

* Try removing addons from Travis

* Fix munge-make-log script

* Try debugging the gcc version

* Remove bulldozer kernel from Linux

* Add c99 arg to setup.py

* Fix addition of -std=c99 flag
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants