Re: [PATCH 4.20 41/65] Revert "powerpc/tm: Unset MSR[TS] if not recheckpointing"

[Michael Ellerman]

Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx> writes:
> On Sat, Jan 12, 2019 at 10:35:59PM +0100, Christoph Biedl wrote:
>> Greg Kroah-Hartman wrote...
>> 
>> > 4.20-stable review patch.  If anyone has any objections, please let me know.
>> >
>> > ------------------
>> >
>> > From: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
>> >
>> > This reverts commit d412deb85a4aada382352a8202beb7af8921cd53 which is
>> > commit 6f5b9f018f4c7686fd944d920209d1382d320e4e upstream.
>> >
>> > It breaks the powerpc build, so drop it from the tree until a fix goes
>> > upstream.
>> 
>> Is this necessary on 4.20? The build failures I reported were on 4.19
>> only. The 4.20.2-rc1 kernel for my Powermac G5 builds with and without
>> that patch, both boot fine, no visible differences. Again however, Breno
>> is authoritative here.
>
> If there's no difference on 4.20, then maybe it's not needed there?  :)
>
> And yes, I would like confirmation from Breno as well.

You shouldn't need the revert on 4.20.

In 4.20 we changed how MSR_TM_ACTIVE() is defined, which means commit
6f5b9f018f4c ("powerpc/tm: Unset MSR[TS] if not recheckpointing") should
build fine on 4.20.

For 4.19 and earlier MSR_TM_ACTIVE() is different and that's what's
causing the build error.

I have a fix queued in my fixes tree and will push it to Linus in the
next few days.

cheers