Generally Mesa updates are keyworded, not masked. It's been 3 releases and I can't find any information regarding the reason for the mask. Reproducible: Always Steps to Reproduce: 1. emerge --sync 2. emerge =mesa-24.3.2 3. cry Actual Results: Masked for testing by Matt Turner Expected Results: Would you like to continue? (Y/n)
The reason is just "masked for testing": ``` # Matt Turner <mattst88@gentoo.org> (2024-11-16) # Masked for testing =dev-util/intel_clc-24.3* =media-libs/mesa-24.3* ``` It was added in: commit 69363f8db3ded5cdcee1ec91980e4e91f40d6eac Author: Matt Turner <mattst88@gentoo.org> Date: Wed Nov 13 19:50:28 2024 -0500 media-libs/mesa: Version bump to 24.3.0_rc2 Signed-off-by: Matt Turner <mattst88@gentoo.org> I'm not sure if this is really worthy of a bug unless you've: a) unmasked it and found it's fine after some daily use (and you're saying "I've provided that testing"), or b) you think it's been far too long masked even without having done any testing of it (I would argue that since mesa-24.3.0 was added on Nov 21st, it's not been *that* long to be unreasonable) Anyway, we'll see what Matt says. (In reply to Tsukasa from comment #0) > Generally Mesa updates are keyworded, not masked. I don't think that means much. Matt is involved (heavily) upstream and if he's either changed the ebuild significantly or aware of significant upstream changes in a release cycle, it's entirely reasonable to add it masked if he thinks it's likely there'll be issues. What generally happens doesn't really change if it's right or wrong to mask a new release by itself.
To a) I've never built anything that required removing a mask before, I assumed they were known to have issues, usually build system related/tool chain all needed to be updated at once. To b) I'm just not used to mesa being masked, I was really just hoping for more information about the why since I'm used to running the newest keyworded versions as they come out. Masking is far more severe from my point of view, so if there were known issues holding it up from general testing like failure's for certain hardware, USE flag issues, etc. I guess filing a bug seems like a complaint, I didn't mean it to be, I just couldn't find more information anywhere. If keyworded ebuilds are deemed good enough for general testing being considered safe I'll work on adjusting my view points. Thank you very much for the rapid response
No worries (and no offence taken at all) -- was just trying to give detail on both our perspective, and also how it's sometimes not clear what people are really asking. (I should say: it's also very possible that Matt masked it just while in RC phase, and forgot to remove it, I don't know.) Let's see what Matt reckons.
Mesa 24.3 enables conformant Vulkan support for AMD GCN1 GPUs: https://www.phoronix.com/news/RADV-GFX6-GFX7-Vulkan-1.3
The bug has been closed via the following commit(s): https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=05ce9a37e8c4b06e413fb4937fc8161d57f2cd03 commit 05ce9a37e8c4b06e413fb4937fc8161d57f2cd03 Author: Matt Turner <mattst88@gentoo.org> AuthorDate: 2024-12-23 14:27:49 +0000 Commit: Matt Turner <mattst88@gentoo.org> CommitDate: 2024-12-23 14:28:13 +0000 profiles: Unmask media-libs/mesa >= 24.3.0 Closes: https://bugs.gentoo.org/946850 Signed-off-by: Matt Turner <mattst88@gentoo.org> profiles/package.mask | 5 ----- 1 file changed, 5 deletions(-)
Yep, just forgot. Thanks for the reminder!
