[PATCH] x86/tsc: Fix early boot crash with bare tsc_early_khz parameter

[PATCH] x86/tsc: Fix early boot crash with bare tsc_early_khz parameter

[Thorsten Blum]

When a bare tsc_early_khz parameter is specified on the kernel command
line without the '=' separator, early parameter parsing passes NULL to
tsc_early_khz_setup(), which dereferences it in kstrtouint() and can
crash the system during early boot.

Reject a NULL argument in tsc_early_khz_setup() by returning -EINVAL so
early parameter parsing reports the malformed option instead.

Fixes: bd35c77e32e4 ("x86/tsc: Add tsc_early_khz command line parameter")
Cc: stable@vger.kernel.org
Signed-off-by: Thorsten Blum <blum@kernel.org>
---
 arch/x86/kernel/tsc.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c
index 723347e2cf7f..82ff4b7490a0 100644
--- a/arch/x86/kernel/tsc.c
+++ b/arch/x86/kernel/tsc.c
@@ -70,6 +70,9 @@ static DEFINE_PER_CPU_ALIGNED(struct cyc2ns, cyc2ns);
 
 static int __init tsc_early_khz_setup(char *buf)
 {
+	if (!buf)
+		return -EINVAL;
+
 	return kstrtouint(buf, 0, &tsc_early_khz);
 }
 early_param("tsc_early_khz", tsc_early_khz_setup);

Re: [PATCH] x86/tsc: Fix early boot crash with bare tsc_early_khz parameter

[Dave Hansen]

On 11/10/26 07:05, Thorsten Blum wrote:
>  static int __init tsc_early_khz_setup(char *buf)
>  {
> +	if (!buf)
> +		return -EINVAL;
> +
>  	return kstrtouint(buf, 0, &tsc_early_khz);
>  }

This _seems_ like the kind of thing that should ideally just be done in
kstrtouint().

Is there a reason we can't just do what kstrtobool() does when handed
buf=NULL?