Re: [PATCH 3.18 14/83] ARM: make lookup_processor_type() non-__init

[Sasha Levin]

On Sun, Dec 02, 2018 at 09:11:50PM +0100, Greg Kroah-Hartman wrote:
> On Sun, Dec 02, 2018 at 11:25:32AM -0500, Sasha Levin wrote:
> > On Sun, Dec 02, 2018 at 05:15:35PM +0100, Greg Kroah-Hartman wrote:
> > > On Sun, Dec 02, 2018 at 10:17:32AM -0500, Sasha Levin wrote:
> > > > On Fri, Nov 30, 2018 at 04:18:43PM +0100, Greg Kroah-Hartman wrote:
> > > > > On Fri, Nov 30, 2018 at 04:15:54PM +0100, Greg Kroah-Hartman wrote:
> > > > > > On Thu, Nov 29, 2018 at 02:28:18PM +0000, Russell King - ARM Linux wrote:
> > > > > > > Hi,
> > > > > > >
> > > > > > > As I've already fed back to Sascha about this, this patch on its own
> > > > > > > does not fix anything, and is not a stable kernel candidate without
> > > > > > > a patch that makes use of it (iow, the spectre fixes.)  It is a
> > > > > > > preparatory patch for mainline commit 383fb3ee8024.
> > > > > > >
> > > > > > > Every commit in:
> > > > > > >
> > > > > > > $ git rev-list v4.16..383fb3ee8024
> > > > > > >
> > > > > > > are the ARM spectre fixes, which are being back-ported by David Long.
> > > > > > >
> > > > > > > Please do not cherry-pick commits from within this series for _any_
> > > > > > > stable kernel, but please wait for David to send you the back-ported
> > > > > > > patches.
> > > > > >
> > > > > > Ugh, ok, that's a bunch here.
> > > > >
> > > > > Ok, not that bad, only 4:
> > > > > 	383fb3ee8024 ("ARM: spectre-v2: per-CPU vtables to work around big.Little systems")
> > > > > 	e209950fdd06 ("ARM: add PROC_VTABLE and PROC_TABLE macros")
> > > > > 	945aceb1db88 ("ARM: clean up per-processor check_bugs method call")
> > > > > 	899a42f83667 ("ARM: make lookup_processor_type() non-__init")
> > > > >
> > > > > I'll go drop them all from the trees now.  Turns out for 4.14 one of
> > > > > these patches broke the build anyway.  Sasha, how did that past your
> > > > > builder tests?
> > > >
> > > > Because it builds :)
> > >
> > > Really?  It broke Guenter's builds...
> >
> > I might be missing something, but Guenter's reply to this -rc thread
> > (https://www.spinics.net/lists/stable/msg272294.html) was:
> >
> > > For v3.18.127-83-gc3953c334bf6:
> > >
> > > Build results:
> > > 	total: 160 pass: 160 fail: 0
> > > Qemu test results:
> > > 	total: 223 pass: 223 fail: 0
>
> The breakage happened on the 4.14 tree with one of the patches above,
> which caused me to do a -rc2.

Oh, I see :(

I suspect that I have a bug somewhere in my scripts that deal with
removing patches from the middle of git branches, I'll try a few
testcases tomorrow.

--
Thanks,
Sasha