Re: [PATCH v5 09/21] KVM: selftests: Add helpers to print proc IRQ affinity for IRQ test

[Sean Christopherson <seanjc@google.com>]

On Thu, Jun 04, 2026, sashiko-bot@kernel.org wrote:
> > +static void print_proc_irq_file(unsigned int irq, const char *file)
> > +{
> > +	char path[PATH_MAX];
> > +	char buf[256];
> 
> [Severity: Medium]
> Will fgets() silently truncate the smp_affinity masks on large systems when
> reading into this buffer?
> 
> CPU affinity masks are formatted as comma-separated 32-bit hex words. A
> 256-byte buffer limits the output to roughly 910 CPUs. 
> 
> Because masks are printed from highest index to lowest, truncation would
> drop the affinity data for lower CPU indices, which are typically where 
> IRQs are pinned.
> 
> > +	FILE *fp;
> > +
> > +	snprintf(path, sizeof(path), "/proc/irq/%u/%s", irq, file);
> > +	fp = fopen(path, "r");
> > +	if (!fp) {
> > +		printf("  Failed to open %s\n", path);
> 
> [Severity: Medium]
> Does this bypass the selftest logging framework?
> 
> KVM selftests typically use pr_info() from test_util.h to allow standard
> informational output to be suppressed when QUIET is defined. Using printf()
> directly forces unconditional output and can cause log spam in CI
> environments.
> 
> > +		return;
> > +	}
> > +
> > +	if (fgets(buf, sizeof(buf), fp)) {
> > +		buf[strcspn(buf, "\n")] = 0;
> > +		printf("  %s: %s\n", path, buf);
> 
> [Severity: Medium]
> Should this also be updated to use pr_info() instead of printf() to respect
> the QUIET mode flag?

I think for the initial landing, it makes sense to skip this and the usage in
irq_test.  Printing the affinity information definitely qualifies as "nice to
have", i.e. it can wait.