linux-kernel - Re: [PATCH] drm/amd/pm: avoid large variable on kernel stack

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]

Message-ID: <CADnq5_P29pHvKPgqvpSMqryOOq6N15=+74vzcDhh47an_ibv2g@mail.gmail.com>
Date:   Fri, 16 Dec 2022 13:24:01 -0500
From:   Alex Deucher <alexdeucher@...il.com>
To:     Christophe JAILLET <christophe.jaillet@...adoo.fr>
Cc:     arnd@...nel.org, llvm@...ts.linux.dev, Jack.Gui@....com,
        KevinYang.Wang@....com, kenneth.feng@....com, Xinhui.Pan@....com,
        ndesaulniers@...gle.com, linux-kernel@...r.kernel.org,
        amd-gfx@...ts.freedesktop.org, nathan@...nel.org, arnd@...db.de,
        dri-devel@...ts.freedesktop.org, trix@...hat.com,
        alexander.deucher@....com, evan.quan@....com,
        christian.koenig@....com, Hawking.Zhang@....com
Subject: Re: [PATCH] drm/amd/pm: avoid large variable on kernel stack

On Thu, Dec 15, 2022 at 2:46 PM Christophe JAILLET
<christophe.jaillet@...adoo.fr> wrote:
>
> Le 15/12/2022 à 17:36, Arnd Bergmann a écrit :
> > From: Arnd Bergmann <arnd-r2nGTMty4D4@...lic.gmane.org>
> >
> > The activity_monitor_external[] array is too big to fit on the
> > kernel stack, resulting in this warning with clang:
> >
> > drivers/gpu/drm/amd/amdgpu/../pm/swsmu/smu13/smu_v13_0_7_ppt.c:1438:12: error: stack frame size (1040) exceeds limit (1024) in 'smu_v13_0_7_get_power_profile_mode' [-Werror,-Wframe-larger-than]
> >
> > Use dynamic allocation instead. It should also be possible to
> > have single element here instead of the array, but this seems
> > easier.
> >
> > Fixes: 334682ae8151 ("drm/amd/pm: enable workload type change on smu_v13_0_7")
> > Signed-off-by: Arnd Bergmann <arnd-r2nGTMty4D4@...lic.gmane.org>
> > ---
> >   .../drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c  | 21 ++++++++++++++-----
> >   1 file changed, 16 insertions(+), 5 deletions(-)
> >
> > diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c
> > index c270f94a1b86..7eba854e09ec 100644
> > --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c
> > +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c
> > @@ -1439,7 +1439,7 @@ static int smu_v13_0_7_get_power_limit(struct smu_context *smu,
> >
> >   static int smu_v13_0_7_get_power_profile_mode(struct smu_context *smu, char *buf)
> >   {
> > -     DpmActivityMonitorCoeffIntExternal_t activity_monitor_external[PP_SMC_POWER_PROFILE_COUNT];
> > +     DpmActivityMonitorCoeffIntExternal_t *activity_monitor_external;
> >       uint32_t i, j, size = 0;
> >       int16_t workload_type = 0;
> >       int result = 0;
> > @@ -1447,6 +1447,12 @@ static int smu_v13_0_7_get_power_profile_mode(struct smu_context *smu, char *buf
> >       if (!buf)
> >               return -EINVAL;
> >
> > +     activity_monitor_external = kcalloc(sizeof(activity_monitor_external),
>
> Hi,
>
> Before, 'activity_monitor_external' was not initialized.
> Maybe kcalloc() is enough?
>
> sizeof(*activity_monitor_external)?
>       ~~~~

I've fixed this up when applying.

Alex

>
> > +                                         PP_SMC_POWER_PROFILE_COUNT,
> > +                                         GFP_KERNEL);
> > +     if (!activity_monitor_external)
> > +             return -ENOMEM;
> > +
> >       size += sysfs_emit_at(buf, size, "                              ");
> >       for (i = 0; i <= PP_SMC_POWER_PROFILE_WINDOW3D; i++)
>
> Unrelated, but wouldn't it be more straightforward with "<
> PP_SMC_POWER_PROFILE_COUNT"?
>
> >               size += sysfs_emit_at(buf, size, "%-14s%s", amdgpu_pp_profile_name[i],
> > @@ -1459,15 +1465,17 @@ static int smu_v13_0_7_get_power_profile_mode(struct smu_context *smu, char *buf
> >               workload_type = smu_cmn_to_asic_specific_index(smu,
> >                                                              CMN2ASIC_MAPPING_WORKLOAD,
> >                                                              i);
> > -             if (workload_type < 0)
> > -                     return -EINVAL;
> > +             if (workload_type < 0) {
> > +                     result = -EINVAL;
> > +                     goto out;
> > +             }
> >
> >               result = smu_cmn_update_table(smu,
> >                                         SMU_TABLE_ACTIVITY_MONITOR_COEFF, workload_type,
> >                                         (void *)(&activity_monitor_external[i]), false);
> >               if (result) {
> >                       dev_err(smu->adev->dev, "[%s] Failed to get activity monitor!", __func__);
> > -                     return result;
> > +                     goto out;
> >               }
> >       }
> >
> > @@ -1495,7 +1503,10 @@ do {                                                                                                   \
> >       PRINT_DPM_MONITOR(Fclk_BoosterFreq);
> >   #undef PRINT_DPM_MONITOR
> >
> > -     return size;
> > +     result = size;
> > +out:
> > +     kfree(activity_monitor_external);
> > +     return result;
> >   }
> >
> >   static int smu_v13_0_7_set_power_profile_mode(struct smu_context *smu, long *input, uint32_t size)
>