linux-kernel - Re: linux-next: build failure after merge of the arm-perf tree

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Date:   Wed, 13 Dec 2023 10:56:23 +0000
From:   Will Deacon <will@...nel.org>
To:     James Clark <james.clark@....com>
Cc:     Stephen Rothwell <sfr@...b.auug.org.au>,
        Catalin Marinas <catalin.marinas@....com>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        Linux Next Mailing List <linux-next@...r.kernel.org>
Subject: Re: linux-next: build failure after merge of the arm-perf tree

On Wed, Dec 13, 2023 at 10:15:25AM +0000, James Clark wrote:
> On 13/12/2023 09:30, Will Deacon wrote:
> > On Wed, Dec 13, 2023 at 10:09:31AM +1100, Stephen Rothwell wrote:
> >> After merging the arm-perf tree, today's linux-next build (x86_64
> >> allmodconfig) failed like this:
> >>
> >> In file included from drivers/perf/arm_dmc620_pmu.c:26:
> >> include/linux/perf/arm_pmu.h:15:10: fatal error: asm/cputype.h: No such file or directory
> >>    15 | #include <asm/cputype.h>
> >>       |          ^~~~~~~~~~~~~~~
> > 
> > Damn, I tested with allnoconfig, defconfig and allmodconfig before I pushed,
> > but only for ARCH=arm64. Thanks for reporting this.
> > 
> >> I can't easily tell what caused this - possibly commit
> >>
> >>   a5f4ca68f348 ("perf/arm_dmc620: Remove duplicate format attribute #defines")
> >>
> >> I have used the arm-perf and arm64 trees from next-20231212 for today
> >> (since the arm64 tree merged the arm-perf tree).
> > 
> > That looks likely. I'll try reverting it and see if it resolves the problem
> > locally.
> > 
> 
> Oops sorry about that, I didn't expect that change to break the x86
> build so I didn't test it there, but it's because of CONFIG_COMPILE_TEST.

No problem, these things happen!

> a5f4ca68f348 is the offending commit, and reverting it would fix it (it
> was just a tidyup). But this change below also works. That include was
> to build ARM_PMU_PROBE which is only used in perf_event_v6.c (and v7)
> which both already include asm/cputype.h. Although maybe that would be a
> bit of a hack.
> 
> diff --git a/include/linux/perf/arm_pmu.h b/include/linux/perf/arm_pmu.h
> index 337f01674b38..a4876b1c481a 100644
> --- a/include/linux/perf/arm_pmu.h
> +++ b/include/linux/perf/arm_pmu.h
> @@ -12,7 +12,6 @@
>  #include <linux/perf_event.h>
>  #include <linux/platform_device.h>
>  #include <linux/sysfs.h>
> -#include <asm/cputype.h>
> 
>  #ifdef CONFIG_ARM_PMU

I agree that this feels like a bit of a bodge, so I'll revert for now and
maybe we can revisit things later.

Cheers,

Will