Re: [PATCH v3 2/2] bootconfig: Fix integer overflow in initrd size check

[Masami Hiramatsu (Google) <mhiramat@kernel.org>]

On Thu, 10 Sep 2026 02:09:01 +0000
sashiko-bot@kernel.org wrote:

> Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
> - [Medium] Mismatched boundary checks for XBC_DATA_MAX cause inconsistent fallback behavior for embedded bootconfigs.
> - [Low] Misleading error message when bootconfig size exceeds XBC_DATA_MAX.
> --
> 
> commit a24325c6fe0374f1cfc3fc8cbe63efbad0f22f32
> Author: Masami Hiramatsu (Google) <mhiramat@kernel.org>
> 
> bootconfig: Fix integer overflow in initrd size check
> 
> This commit fixes an integer overflow where a crafted initrd with a huge
> bootconfig size can cause pointer arithmetic wrap-around. It ensures the
> initrd bounds are respected and verifies the size does not exceed max limits.
> 
> Closes: https://lore.kernel.org/all/20260910010137.EE0431F000FF@smtp.kernel.org/
> 
> > diff --git a/init/main.c b/init/main.c
> > index a9362b2cc258a..de24c41ce815e 100644
> > --- a/init/main.c
> > +++ b/init/main.c
> 
> [ ... ]
> 
> > @@ -294,16 +295,21 @@ static void * __init get_boot_config_from_initrd(size_t *_size)
> >  
> >  found:
> >  	hdr = (u8 *)(data - 8);
> > +	if ((unsigned long)hdr < initrd_start)
> > +		return NULL;
> > +
> >  	size = get_unaligned_le32(hdr);
> >  	csum = get_unaligned_le32(hdr + 4);
> >  
> > -	data = ((void *)hdr) - size;
> > -	if ((unsigned long)data < initrd_start) {
> > -		pr_err("bootconfig size %d is greater than initrd size %ld\n",
> > +	if (size > XBC_DATA_MAX ||
> 
> [Severity: Medium]
> If the size is exactly equal to XBC_DATA_MAX, this check passes and the
> function returns a valid data pointer. This prevents the caller from falling
> back to the embedded bootconfig:
> 
> init/main.c:setup_boot_config() {
> 	...
> 	data = get_boot_config_from_initrd(&size);
> 	if (!data) {
> 		data = xbc_get_embedded_bootconfig(&size);
> 		...
> 	}
> 	...
> }
> 
> However, setup_boot_config() uses inclusive inequality for XBC_DATA_MAX:
> 
> init/main.c:setup_boot_config() {
> 	...
> 	if (size >= XBC_DATA_MAX) {
> 		pr_err("bootconfig size %ld greater than max size %d\n", ...);
> 		return;
> 	}
> 	...
> }

Hmm, this is wrong. Since XBC_DATA_MAX is including maximum size, 
setup_boot_config() should change this condition.

> 
> Because the limits don't match, an exact size of XBC_DATA_MAX causes
> setup_boot_config() to reject the payload and abort without attempting
> the fallback. Should this check use size >= XBC_DATA_MAX to match?
> 
> > +	    size > ((unsigned long)hdr - initrd_start)) {
> > +		pr_err("bootconfig size %u is greater than initrd size %lu\n",
> > +			size, initrd_end - initrd_start);
> 
> [Severity: Low]
> If this error path is taken specifically because size > XBC_DATA_MAX, the
> error message claims the bootconfig size is greater than the initrd size.
> Can this message be updated to accurately reflect which limit was exceeded?

Ah, good catch :)

Thanks,

> 
> >  		return NULL;
> >  	}
> >  
> > +	data = ((void *)hdr) - size;
> > +
> >  	if (xbc_calc_checksum(data, size) != csum) {
> >  		pr_err("bootconfig checksum failed\n");
> >  		return NULL;
> 
> -- 
> Sashiko AI review · https://sashiko.dev/#/patchset/178900555453.200943.4719889510443114485.stgit@devnote2?part=2


-- 
Masami Hiramatsu (Google) <mhiramat@kernel.org>