Skip to contentSkip to conversation

x86, microcode: Move to a proper location

tip:x86/microcode H. Peter Anvin 3 participants 9 messages Jan 14, 2014 1 message not in the archive

lore
Share & tools
lore

A message that is not in the archive (sent off-list or not yet received): [email protected]

#1 lore raw
Quoted 9 lines · tip-bot for Borislav Petkov
Commit-ID:  bad5fa631fca5466401cd4a48e30cc1f1cb6101e
Gitweb:     http://git.kernel.org/tip/bad5fa631fca5466401cd4a48e30cc1f1cb6101e
Author:     Borislav Petkov <[email protected]>
AuthorDate: Sun, 1 Dec 2013 18:09:58 +0100
Committer:  Borislav Petkov <[email protected]>
CommitDate: Mon, 13 Jan 2014 20:00:12 +0100

x86, microcode: Move to a proper location

Quoted 8 lines

We've grown a bunch of microcode loader files all prefixed with "microcode_". They should be under cpu/ because this is strictly CPU-related functionality so do that and drop the prefix since they're in their own directory now which gives that prefix. :)

While at it, drop MICROCODE_INTEL_LIB config item and stash the functionality under CONFIG_MICROCODE_INTEL as it was its only user.

Quite frankly I would be much happier if we didn't stash so much under arch/x86/kernel/cpu ... quite frankly it feels like almost *anything* could go under there. The microcode code, for example, could go under its own subtree.

Both kernel and kernel/cpu really could use a house cleaning and actually separate things out into better categories and avoid needlessly deep pathnames.

-hpa

#2 lore raw
Quoted 19 lines · H. Peter Anvin
Quoted 9 lines · tip-bot for Borislav Petkov
Commit-ID:  bad5fa631fca5466401cd4a48e30cc1f1cb6101e
Gitweb:     http://git.kernel.org/tip/bad5fa631fca5466401cd4a48e30cc1f1cb6101e
Author:     Borislav Petkov <[email protected]>
AuthorDate: Sun, 1 Dec 2013 18:09:58 +0100
Committer:  Borislav Petkov <[email protected]>
CommitDate: Mon, 13 Jan 2014 20:00:12 +0100

x86, microcode: Move to a proper location

Quoted 8 lines

We've grown a bunch of microcode loader files all prefixed with "microcode_". They should be under cpu/ because this is strictly CPU-related functionality so do that and drop the prefix since they're in their own directory now which gives that prefix. :)

While at it, drop MICROCODE_INTEL_LIB config item and stash the functionality under CONFIG_MICROCODE_INTEL as it was its only user.

Quoted 8 lines

Quite frankly I would be much happier if we didn't stash so much under arch/x86/kernel/cpu ... quite frankly it feels like almost *anything* could go under there. The microcode code, for example, could go under its own subtree.

Both kernel and kernel/cpu really could use a house cleaning and actually separate things out into better categories and avoid needlessly deep pathnames.

Absolutely. I think moving arch/x86/kernel/cpu/ to arch/x86/cpu/ would be a first good step. It's all kernel code, there's other subdirectories there that are kernel code ('pci/', 'platform/', etc.), so there's little point in continuing the historic accident of a 'kernel/' subdirectory.

Once that is done we can move certain things outside as well - for example arch/x86/cpu/perf/ could probably move to arch/x86/events/, because that code is not about CPU events anymore either.

Thanks,

Ingo

#3 lore raw
Quoted 9 lines · Ingo Molnar

Absolutely. I think moving arch/x86/kernel/cpu/ to arch/x86/cpu/ would be a first good step. It's all kernel code, there's other subdirectories there that are kernel code ('pci/', 'platform/', etc.), so there's little point in continuing the historic accident of a 'kernel/' subdirectory.

Once that is done we can move certain things outside as well - for example arch/x86/cpu/perf/ could probably move to arch/x86/events/, because that code is not about CPU events anymore either.

All fine and dandy with me, but I'd prefer to do this after the merge window, i.e. for 3.15. Rushing it now would just cause unnecessary trouble IMO.

Signature · 5 lines
Regards/Gruss,
    Boris.

Sent from a fat crate under my desk. Formatting is fine.
--
#4 lore raw
Quoted 7 lines · Borislav Petkov
Quoted 6 lines · Ingo Molnar

Absolutely. I think moving arch/x86/kernel/cpu/ to arch/x86/cpu/ would be a first good step. It's all kernel code, there's other subdirectories there that are kernel code ('pci/', 'platform/', etc.), so there's little point in continuing the historic accident of a 'kernel/' subdirectory.

Quoted 8 lines
Quoted 3 lines

Once that is done we can move certain things outside as well - for example arch/x86/cpu/perf/ could probably move to arch/x86/events/, because that code is not about CPU events anymore either.

All fine and dandy with me, but I'd prefer to do this after the merge window, i.e. for 3.15. Rushing it now would just cause unnecessary trouble IMO.

Oh, absolutely.

-hpa

#5 lore raw
Quoted 12 lines · H. Peter Anvin
Quoted 11 lines · Borislav Petkov
Quoted 9 lines · Ingo Molnar

Absolutely. I think moving arch/x86/kernel/cpu/ to arch/x86/cpu/ would be a first good step. It's all kernel code, there's other subdirectories there that are kernel code ('pci/', 'platform/', etc.), so there's little point in continuing the historic accident of a 'kernel/' subdirectory.

Once that is done we can move certain things outside as well - for example arch/x86/cpu/perf/ could probably move to arch/x86/events/, because that code is not about CPU events anymore either.

Quoted 6 lines
Quoted 4 lines

All fine and dandy with me, but I'd prefer to do this after the merge window, i.e. for 3.15. Rushing it now would just cause unnecessary trouble IMO.

Oh, absolutely.

Agreed.

Thanks,

Ingo

#6 lore raw
In reply to H. Peter Anvin
Quoted 4 lines · H. Peter Anvin

Quite frankly I would be much happier if we didn't stash so much under arch/x86/kernel/cpu ... quite frankly it feels like almost *anything* could go under there. The microcode code, for example, could go under its own subtree.

Yeah, about the microcode, shouldn't it be under kernel/cpu/? My train of thought is: because it is cpu-related functionality, it logically belongs there.

Or do we rather apply the shallow tree logic of

arch/x86/microcode
arch/x86/events
...

?

I have to say, arch/x86/kernel/cpu/ is kinda long and hard to remember so the shallow structure should be nicer on the eyes and the brain :)

Signature · 5 lines
Regards/Gruss,
    Boris.

Sent from a fat crate under my desk. Formatting is fine.
--
#7 lore raw
Quoted 9 lines · Borislav Petkov
Quoted 4 lines · H. Peter Anvin

Quite frankly I would be much happier if we didn't stash so much under arch/x86/kernel/cpu ... quite frankly it feels like almost *anything* could go under there. The microcode code, for example, could go under its own subtree.

Yeah, about the microcode, shouldn't it be under kernel/cpu/? My train of thought is: because it is cpu-related functionality, it logically belongs there.

arch/x86/cpu/microcode seems clean. I'm wondering if what is mostly currently in arch/x86/kernel/cpu should be mostly in something like arch/x86/cpu/init.

-hpa

#8 lore raw
Quoted 3 lines · H. Peter Anvin

arch/x86/cpu/microcode seems clean. I'm wondering if what is mostly currently in arch/x86/kernel/cpu should be mostly in something like arch/x86/cpu/init.

Yeah, I can start reorganizing stuff slowly and with time, the best place will dawn on us. One thing's for sure, the "kernel" in the path is kinda redundant because, as Ingo said, it is all kernel.

Signature · 5 lines
Regards/Gruss,
    Boris.

Sent from a fat crate under my desk. Formatting is fine.
--
#9 lore raw
Quoted 9 lines · Borislav Petkov
Quoted 3 lines · H. Peter Anvin

arch/x86/cpu/microcode seems clean. I'm wondering if what is mostly currently in arch/x86/kernel/cpu should be mostly in something like arch/x86/cpu/init.

Yeah, I can start reorganizing stuff slowly and with time, the best place will dawn on us. One thing's for sure, the "kernel" in the path is kinda redundant because, as Ingo said, it is all kernel.

Indeed.

-hpa