ci: fix the ctest sanitize runs

[netrunnereve]

Those jobs are getting stuck as we only have three machines with the cpu tag.

Requirements

• I have read and agree with the contributing guidelines
• AI usage disclosure: NO

[netrunnereve]

Aaaand it crashed, back to draft. Actually the thread test seems to have crashed sometimes in the past as well.

[netrunnereve]

From looking at the old jobs it looks like the thread tests only pass on ggml-6-x86-vulkan-t4 and ggml-8-x86-nvidia-a10 while it crashes on sg-hl1-ci-nvidia-vulkan-cm. It crashes on my own computer as well with the same ThreadSanitizer: unexpected memory mapping error and it's probably due to how ASLR is setup on that particular machine.

So right now I've made that particular test only run on those two specific machines and we can always add more to the list or have a special tag for it in the future. Or maybe this test can be straight up removed altogether, I dunno.

[netrunnereve]

Okay I think it makes more sense to just run the thread test on the Github machines instead of hardcoding a possibly changing list of self hosted machines to run it on.

[CISC]

Strange, this started failing for no reason:
https://github.com/ggml-org/llama.cpp/actions/runs/31362342045/job/93373543781#step:7:30

[netrunnereve]

Strange, this started failing for no reason: https://github.com/ggml-org/llama.cpp/actions/runs/31362342045/job/93373543781#step:7:30

Ugh I think it's something to do with the Intel-LNL-U7-258V machine, I checked the jobs after my pr got merged and all the ones that failed were on that machine. I'll deal with this later.