Re: [PATCH v2] x86: disable non-instrumented version of copy_mc when KMSAN is enabled

[Dave Hansen]

On 3/1/24 14:52, Tetsuo Handa wrote:
>  unsigned long __must_check copy_mc_to_kernel(void *dst, const void *src, unsigned len)
>  {
> -	if (copy_mc_fragile_enabled)
> +	if (!IS_ENABLED(CONFIG_KMSAN) && copy_mc_fragile_enabled)
>  		return copy_mc_fragile(dst, src, len);
> -	if (static_cpu_has(X86_FEATURE_ERMS))
> +	if (!IS_ENABLED(CONFIG_KMSAN) && static_cpu_has(X86_FEATURE_ERMS))
>  		return copy_mc_enhanced_fast_string(dst, src, len);
>  	memcpy(dst, src, len);
>  	return 0;
> @@ -74,14 +74,14 @@ unsigned long __must_check copy_mc_to_user(void __user *dst, const void *src, un
>  {
>  	unsigned long ret;
>  
> -	if (copy_mc_fragile_enabled) {
> +	if (!IS_ENABLED(CONFIG_KMSAN) && copy_mc_fragile_enabled) {
>  		__uaccess_begin();
>  		ret = copy_mc_fragile((__force void *)dst, src, len);
>  		__uaccess_end();
>  		return ret;
>  	}
>  
> -	if (static_cpu_has(X86_FEATURE_ERMS)) {
> +	if (!IS_ENABLED(CONFIG_KMSAN) && static_cpu_has(X86_FEATURE_ERMS)) {
>  		__uaccess_begin();
>  		ret = copy_mc_enhanced_fast_string((__force void *)dst, src, len);
>  		__uaccess_end();

Where does the false positive _come_ from?  Can we fix copy_mc_fragile()
and copy_mc_enhanced_fast_string() instead of just not using them?

The three enable_copy_mc_fragile() are presumably doing so for a reason.
 What is this patch's impact on _those_?

Third, instead of sprinkling IS_ENABLED(CONFIG_KMSAN) checks in random
spots, can we do this in a central spot?