FSGSBASE v12 Patches Published With Even Microsoft Intrigued By Its Performance
Just this weekend we were writing about the eleventh spin of the Linux FSGSBASE patches for this capability that can help out performance going back to Ivy Bridge era processors. Kicking off Monday, version twelve is already out and an explanation why a Microsoft developer has been stewarding these patches across the finish line.
See the aforelinked article for details on this ongoing saga of mainlining Linux support for FSGSBASE... It's been going on for a long time but now at least it looks like it may have the momentum to get it across the finish line. But how did this go from being something Intel developers were sending out patches on to now Microsoft's Linux kernel developer, Sasha Levin, volleying the new patch series?
Longtime Intel Linux developer David Hansen brought up, "I'm also very curious how you came to be submitting this series. Is this a team effort between you, the And[iy]s and Chang? Or, were you just trying to help out? I was hoping to see some acknowledgement of this situation in the cover letter but didn't see anything."
Sasha responded with, "What happened here was that v9 needed to be rebased on top of v5.7 which required some changes. I did the rebase and sent it to Andi and Chang who have suggested that I'll just send it out myself. There was no planning beyond that. My interest in this is that we have a few workloads that value the ability to access FS/GS base directly and show nice performance improvement with this patchset. I'm not a fan of carrying stuff out of tree :)"
So even Microsoft has found some of their own Linux-based workloads benefiting from FSGSBASE and wanting to get the functionality merged. Presumably this is for some Azure workloads but there was no comment beyond the above statement.
In any case, Sasha has now sent out FSGSBASE v12 patches that re-base against the latest Linux Git code and also fix a 32-bit build error.
It will be interesting to see if this work finally gets merged for the upcoming Linux 5.8 cycle. I've been running some fresh FSGSBASE (v11) patches and will have those performance benchmarks published this week.
See the aforelinked article for details on this ongoing saga of mainlining Linux support for FSGSBASE... It's been going on for a long time but now at least it looks like it may have the momentum to get it across the finish line. But how did this go from being something Intel developers were sending out patches on to now Microsoft's Linux kernel developer, Sasha Levin, volleying the new patch series?
Longtime Intel Linux developer David Hansen brought up, "I'm also very curious how you came to be submitting this series. Is this a team effort between you, the And[iy]s and Chang? Or, were you just trying to help out? I was hoping to see some acknowledgement of this situation in the cover letter but didn't see anything."
Sasha responded with, "What happened here was that v9 needed to be rebased on top of v5.7 which required some changes. I did the rebase and sent it to Andi and Chang who have suggested that I'll just send it out myself. There was no planning beyond that. My interest in this is that we have a few workloads that value the ability to access FS/GS base directly and show nice performance improvement with this patchset. I'm not a fan of carrying stuff out of tree :)"
So even Microsoft has found some of their own Linux-based workloads benefiting from FSGSBASE and wanting to get the functionality merged. Presumably this is for some Azure workloads but there was no comment beyond the above statement.
In any case, Sasha has now sent out FSGSBASE v12 patches that re-base against the latest Linux Git code and also fix a 32-bit build error.
It will be interesting to see if this work finally gets merged for the upcoming Linux 5.8 cycle. I've been running some fresh FSGSBASE (v11) patches and will have those performance benchmarks published this week.
1 Comment