Hi William, On Tue, Mar 08, 2022 at 04:06:45PM +0100, William Lallemand wrote: > Hello, > > The attached patch implements a somehow ugly way to cache the VTest > binary, basically it gets the commit ID by doing a curl of the > master.patch on the github URL. > > It allows to save ~8s per matrix row, which is around 160s in total. I > know there is a small window where the curl and the git clone won't have > the same ID but that will be rebuild anyway for the next build, so > that's fine in my opinion. > > We could probably use the same approach to cache quictls or anything > that uses a git repository. > > Also, I'm wondering if we could also cache the build of HAProxy, you > could think that weird, but in fact it will help relaunch the tests when > one is failing, without rebuilding the whole thing. > > Let me know if we can improve the attached patch, otherwise I'll merge > it.
Thanks for this. I can't judge on how this can be improved, however I noticed today that one of the important goals of the CI is to see if stuff builds at all (and we often break one platform or another). Due to this I think we should move the build of vtest after the build of haproxy (and generally, anything that's not needed for the build ought to be moved after). This will at least save whatever can be saved on failed builds. Willy
