Re: [PATCH] cpufreq: amd-pstate-ut: fix null pointer dereference

[K Prateek Nayak <kprateek.nayak@amd.com>]

Hello Qianheng,

On 7/14/2026 1:29 PM, Qianheng Peng wrote:
> The crash issue may occur when insmod amd_pstate_ut modules.
> 
>    amd_pstate_ut: 1    amd_pstate_ut_acpi_cpc_valid  success!
>    amd_pstate_ut: 2    amd_pstate_ut_check_enabled   success!
>    BUG: kernel NULL pointer dereference, address: 0000000000000080
>    #PF: supervisor read access in kernel mode
>    #PF: error_code(0x0000) - not-present page
>    PGD 0 P4D 0
>    Oops: 0000 [#1] SMP NOPTI
>    CPU: 0 PID: 20300 Comm: modprobe
>    Kdump: loaded Tainted: G O 6.6.0-0010.rc1.ctl4.x86_64 #1
>    Hardware name: FiberHome R2200 V5/Xeon Boards, BIOS 3.1a 02/24/2020

Is this an AMD platform?

>    RIP: 0010:amd_pstate_ut_check_perf+0x141/0x280 [amd_pstate_ut]
>    Call Trace:
>     <TASK>
>     amd_pstate_ut_init+0x1b/0xff0 [amd_pstate_ut]
>     ? __pfx_amd_pstate_ut_init+0x10/0x10 [amd_pstate_ut]
>     do_one_initcall+0x42/0x2e0
>     ? kmalloc_trace+0x26/0x90
>     do_init_module+0x60/0x240
>     __se_sys_init_module+0x185/0x1c0
>     do_syscall_64+0x62/0x190
>     entry_SYSCALL_64_after_hwframe+0x76/0x7e
>     </TASK>
> 
> Add invalidation check of cpudata in amd_pstate_ut_check_perf() and
> amd_pstate_ut_check_freq() to avoid unpredicated null pointer dereference.
> 
> Fixes: 14eb1c96e3a3 ("cpufreq: amd-pstate: Add test module for amd-pstate driver")
> Suggested-by: Li Xiong <xiongl24@chinatelecom.cn>
> Suggested-by: Xibo Wang <wangxb12@chinatelecom.cn>
> Signed-off-by: Qianheng Peng <pengqh1@chinatelecom.cn>
> ---
>  drivers/cpufreq/amd-pstate-ut.c | 8 ++++++++
>  1 file changed, 8 insertions(+)
> 
> diff --git a/drivers/cpufreq/amd-pstate-ut.c b/drivers/cpufreq/amd-pstate-ut.c
> index 735b29f..3b24260 100644
> --- a/drivers/cpufreq/amd-pstate-ut.c
> +++ b/drivers/cpufreq/amd-pstate-ut.c
> @@ -157,6 +157,10 @@ static int amd_pstate_ut_check_perf(u32 index)
>                 if (!policy)
>                         continue;
>                 cpudata = policy->driver_data;
> +               if (!cpudata) {
> +                       pr_err("%s driver_data of %s is empty!\n", __func__, policy->kobj.name);
> +                       return -EINVAL;
> +               }
> 
>                 if (get_shared_mem()) {
>                         ret = cppc_get_perf_caps(cpu, &cppc_perf);
> @@ -229,6 +233,10 @@ static int amd_pstate_ut_check_freq(u32 index)
>                 if (!policy)
>                         continue;
>                 cpudata = policy->driver_data;
> +               if (!cpudata) {
> +                       pr_err("%s driver_data of %s is empty!\n", __func__, policy->kobj.name);
> +                       return -EINVAL;
> +               }

The only time I think this can happen is when the amd-pstate-ut races with a
the amd_pstate.*_cpu_exit() (probably as a result of concurrent mode switch)
or if you are running this without running amd-pstate as the cpufreq driver
(which the ut doesn't check for beforehand)

My setup fails at this exact same point when I have the driver unloaded.
What is the output of following before you run into this issue:

    cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_driver
    cat /sys/devices/system/cpu/amd_pstate/status

Are you doing any concurrent hotplug / mode switch?

> 
>                 if (!((policy->cpuinfo.max_freq >= cpudata->nominal_freq) &&
>                         (cpudata->nominal_freq > cpudata->lowest_nonlinear_freq) &&
> --
> 1.8.3.1
> 

-- 
Thanks and Regards,
Prateek