Added travisCI SIMDe example for arm64 + fix to Makefile for SIMDe#604
Added travisCI SIMDe example for arm64 + fix to Makefile for SIMDe#604mbrcic wants to merge 2 commits into
Conversation
|
Thanks! awesome! Let's wait @lh3 's response. |
| language: c | ||
| compiler: gcc | ||
| script: make arm_neon=1 aarch64=1 | ||
| arch: arm64 |
There was a problem hiding this comment.
I think this line's syntax is wrong. You need -.
- arch: arm64
There was a problem hiding this comment.
This is the result of the CI: https://travis-ci.org/github/lh3/minimap2/builds/685858196 . There are total 6 jobs. But it has to be total 7 jobs.
There was a problem hiding this comment.
yeah, sorry, IDE issues :)
There was a problem hiding this comment.
with that syntax mistake, the file is parsed in such a way that arm64 config takes only one (the last) variable values. you can check the parsed values with validator: https://config.travis-ci.com/explore
when that - is put in place, it parses well and there should be 7 tests.
There was a problem hiding this comment.
I did not know the config validation page. Thanks for sharing.
Yeah, now total 7 jobs here. https://travis-ci.org/github/lh3/minimap2/builds/685864104
There was a problem hiding this comment.
If we use name: syntax in .travis.yml, technically we can distinguish the similar jobs by the job names on the CI page. But I assume Heng likes the minimum config, and does not like the name syntax. So, maybe the current config is okay.
|
With this syntax mistake fixed, travisCI should do 7 tests, twice on arm64. |
| PROG= minimap2 | ||
| PROG_EXTRA= sdust minimap2-lite | ||
| LIBS= -lm -lz -lpthread | ||
| SSE4= |
There was a problem hiding this comment.
I found the tailing space (tab) here. It's better to remove it.
There was a problem hiding this comment.
Removed. Should this fixing commit be squashed as well?
|
I think it's better to unify this PR's 2 commits to 1 commit by doing squash. (such as |
Reproductive command steps for arm64 minimap2 using SIMDe, as asked in:
#597
In process, I have fixed an error in Makefile.simde that probably appeared during the merge of previous PR.