From: Scott Wood <scottwood@freescale.com>
To: Michael Ellerman <mpe@ellerman.id.au>
Cc: <linuxppc-dev@lists.ozlabs.org>,
Tiejun Chen <tiejun.chen@intel.com>, <kexec@lists.infradead.org>
Subject: Re: [RFC,14/17] powerpc/book3e-64/kexec: Enable SMP release
Date: Tue, 18 Aug 2015 00:09:19 -0500 [thread overview]
Message-ID: <1439874559.15601.70.camel@freescale.com> (raw)
In-Reply-To: <20150818045141.D1E8C14032F@ozlabs.org>
On Tue, 2015-08-18 at 14:51 +1000, Michael Ellerman wrote:
> On Sat, 2015-18-07 at 20:08:51 UTC, Scott Wood wrote:
> > booted_from_exec is similar to __run_at_load, except that it is set for
> ^
> missing k.
>
> Also do you mind using __booted_from_kexec to keep the naming similar to the
> other variables down there, and also make it clear it's low level guts.
>
> I see you asked for them to be removed on the original patch but all the
> other
> vars in there are named that way.
I'm not a fan of it as it isn't distinguishing from a non-underscore version,
isn't there for namespacing reasons, and it's not even a private
implementation detail -- it's part of the interface with kexec tools. I'll
change it if you want, though.
> > diff --git a/arch/powerpc/kernel/head_64.S b/arch/powerpc/kernel/head_64.S
> > index 1b77956..ae2d6b5 100644
> > --- a/arch/powerpc/kernel/head_64.S
> > +++ b/arch/powerpc/kernel/head_64.S
> > @@ -91,6 +91,21 @@ __secondary_hold_spinloop:
> > __secondary_hold_acknowledge:
> > .llong 0x0
> >
> > + /* Do not move this variable as kexec-tools knows about it. */
> > + . = 0x58
> > + .globl booted_from_kexec
> > +booted_from_kexec:
> > + /*
> > + * "nkxc" -- not (necessarily) from kexec by default
> > + *
> > + * This flag is set to 1 by a loader if the kernel is being
> > + * booted by kexec. Older kexec-tools don't know about this
> > + * flag, so platforms other than fsl-book3e should treat a value
> > + * of "nkxc" as inconclusive. fsl-book3e relies on this to
> > + * know how to release secondary cpus.
> > + */
> > + .long 0x6e6b7863
>
> Couldn't we say that "nkxc" (whatever that stands for)
It stands for "no kexec", which is true if that value is not overwritten.
> means "unknown", and
> have kexec-tools write "yes" to indicate yes. I realise that's not 100%
> bullet
That is what I implemented (other than "1" versus "yes").
> > diff --git a/arch/powerpc/kernel/setup_64.c
> > b/arch/powerpc/kernel/setup_64.c
> > index 505ec2c..baeddcc 100644
> > --- a/arch/powerpc/kernel/setup_64.c
> > +++ b/arch/powerpc/kernel/setup_64.c
> > @@ -340,11 +340,23 @@ void early_setup_secondary(void)
> > #endif /* CONFIG_SMP */
> >
> > #if defined(CONFIG_SMP) || defined(CONFIG_KEXEC)
> > +static bool use_spinloop(void)
> > +{
> > +#ifdef CONFIG_PPC_FSL_BOOK3E
> > + return booted_from_kexec == 1;
> > +#else
> > + return true;
> > +#endif
>
> Ugh, more ifdefs.
>
> What about:
>
> return IS_ENABLED(CONFIG_PPC_FSL_BOOK3E) && (booted_from_kexec == 1);
>
> If that works, I haven't checked. It's slightly less ugly?
That would return "false" for non-book3e which isn't correct.
If it has to be done with a single expression, then it'd be:
return !IS_ENABLED(CONFIG_PPC_FSL_BOOK3E) || booted_from_kexec == 1;
-Scott
next prev parent reply other threads:[~2015-08-18 5:09 UTC|newest]
Thread overview: 25+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-07-18 20:08 [RFC PATCH 00/17] powerpc/fsl-book3e-64: kexec/kdump support Scott Wood
2015-07-18 20:08 ` [RFC PATCH 01/17] powerpc/85xx: Load all early TLB entries at once Scott Wood
2015-07-18 20:08 ` [RFC PATCH 02/17] powerpc/85xx: Don't use generic timebase sync on 64-bit Scott Wood
2015-07-18 20:08 ` [RFC PATCH 03/17] crypto: caam: Blacklist CAAM when kexec is enabled Scott Wood
2015-07-18 20:08 ` [RFC PATCH 04/17] powerpc/fsl-corenet: Disable coreint if " Scott Wood
2015-07-18 20:08 ` [RFC PATCH 05/17] powerpc/fsl-booke-64: Don't limit ppc64_rma_size to one TLB entry Scott Wood
2015-07-18 20:08 ` [RFC PATCH 06/17] powerpc/85xx: Implement 64-bit kexec support Scott Wood
2015-07-18 20:08 ` [RFC PATCH 07/17] powerpc/e6500: kexec: Handle hardware threads Scott Wood
2015-07-18 20:08 ` [RFC PATCH 08/17] powerpc/book3e-64: rename interrupt_end_book3e with __end_interrupts Scott Wood
2015-07-18 20:08 ` [RFC PATCH 09/17] powerpc/booke64: Fix args to copy_and_flush Scott Wood
2015-07-18 20:08 ` [RFC PATCH 10/17] powerpc/book3e: support CONFIG_RELOCATABLE Scott Wood
2015-07-18 20:08 ` [RFC PATCH 11/17] powerpc/book3e/kdump: Enable crash_kexec_wait_realmode Scott Wood
2015-07-18 20:08 ` [RFC PATCH 12/17] powerpc/book3e-64: Don't limit paca to 256 MiB Scott Wood
2015-07-18 20:08 ` [RFC PATCH 13/17] powerpc/book3e-64/kexec: create an identity TLB mapping Scott Wood
2015-07-18 20:08 ` [RFC PATCH 14/17] powerpc/book3e-64/kexec: Enable SMP release Scott Wood
2015-08-18 4:51 ` [RFC,14/17] " Michael Ellerman
2015-08-18 5:09 ` Scott Wood [this message]
2015-08-20 4:54 ` [RFC PATCH 14/17] " Michael Ellerman
2015-08-24 20:25 ` Scott Wood
2015-08-25 1:57 ` Michael Ellerman
2015-08-25 23:40 ` Scott Wood
2015-08-26 1:13 ` Michael Ellerman
2015-07-18 20:08 ` [RFC PATCH 15/17] powerpc/booke: Only use VIRT_PHYS_OFFSET on booke32 Scott Wood
2015-07-18 20:08 ` [RFC PATCH 16/17] powerpc/book3e-64/kexec: Set "r4 = 0" when entering spinloop Scott Wood
2015-07-18 20:08 ` [RFC PATCH 17/17] powerpc/book3e-64: Enable kexec Scott Wood
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1439874559.15601.70.camel@freescale.com \
--to=scottwood@freescale.com \
--cc=kexec@lists.infradead.org \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=mpe@ellerman.id.au \
--cc=tiejun.chen@intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).