Re: [PATCH v8 05/20] KVM: selftests: Seed libc's RNG before using it to generate a seed for KVM's pRNG

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] The KVM pseudo-random number generator gets trapped generating only 0s if `random()` returns 0 or 2147483647.
--

commit cb357af51bf4702a2c21b353556fb23a80cc5efe
Author: Sean Christopherson <seanjc@google.com>

KVM: selftests: Seed libc's RNG before using it to generate a seed for KVM's pRNG

This commit initializes the libc RNG with the current time before using it
to generate the initial seed for KVM's selftest pseudo-random number
generator.

> diff --git a/tools/testing/selftests/kvm/lib/kvm_util.c b/tools/testing/selftests/kvm/lib/kvm_util.c
> index 1016865d3f7a2..277166ab1aa99 100644
> --- a/tools/testing/selftests/kvm/lib/kvm_util.c
> +++ b/tools/testing/selftests/kvm/lib/kvm_util.c

[ ... ]

> @@ -2284,6 +2285,7 @@ void __attribute((constructor)) kvm_selftest_init(void)
>  	sigaction(SIGILL, &sig_sa, NULL);
>  	sigaction(SIGFPE, &sig_sa, NULL);
>  
> +	srandom(time(0));
>  	kvm_seed_rng(random());

[Severity: Medium]
Can this cause the KVM PRNG to get stuck returning only 0s, resulting in
intermittent test failures?

If random() happens to return 0 or 2147483647 (the modulus), passing it
to kvm_seed_rng() will cause the Lehmer LCG calculation in kvm_random_u32()
to permanently evaluate to 0:

tools/testing/selftests/kvm/lib/test_util.c:kvm_random_u32() {
    ...
    state->seed = (u64)state->seed * 48271 % ((u32)(1 << 31) - 1);
    return state->seed;
}

When state->seed becomes 0, all subsequent PRNG calls will endlessly return 0.
Does it make sense to sanitize the result of random() here to avoid these
pathological seeds?

>  
>  	kvm_selftest_arch_init();

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260626213534.3866178-1-seanjc@google.com?part=5