linux-kernel - Re: [PATCH v2 1/3] PCI: Relaxed tail alignment should never increase min

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <21e11870-f125-e9e7-04f3-ade94d6be6b1@linux.intel.com>
Date: Thu, 21 Aug 2025 19:45:17 +0300 (EEST)
From: Ilpo Järvinen <ilpo.jarvinen@...ux.intel.com>
To: Markus Elfring <Markus.Elfring@....de>
cc: linux-pci@...r.kernel.org, Bjorn Helgaas <bhelgaas@...gle.com>, 
    stable@...r.kernel.org, tudor.ambarus@...aro.org, 
    LKML <linux-kernel@...r.kernel.org>, kernel-janitors@...r.kernel.org, 
    Christian König <christian.koenig@....com>, 
    D Scott Phillips <scott@...amperecomputing.com>, 
    Krzysztof Wilczyński <kw@...ux.com>, 
    Rio Liu <rio@....me>
Subject: Re: [PATCH v2 1/3] PCI: Relaxed tail alignment should never increase
 min_align

On Thu, 21 Aug 2025, Markus Elfring wrote:

> …
> > Ensure min_align is not increased by the relaxed tail alignment.
> …
> 
> 
> …
> +++ b/drivers/pci/setup-bus.c
> …
> @@ -1261,8 +1263,9 @@ static int pbus_size_mem(struct pci_bus *bus, unsigned long mask,
>  		if (bus->self && size1 &&
>  		    !pbus_upstream_space_available(bus, mask | IORESOURCE_PREFETCH, type,
>  						   size1, add_align)) {
> -			min_align = 1ULL << (max_order + __ffs(SZ_1M));
> -			min_align = max(min_align, win_align);
> +			relaxed_align = 1ULL << (max_order + __ffs(SZ_1M));
> +			relaxed_align = max(min_align, win_align);
> …
> 
> I wonder why a variable content would be overwritten here
> without using the previous value.
> https://cwe.mitre.org/data/definitions/563.html

Hi Markus,

This looks a very good catch. I think it too should have been:

relaxed_align = max(relaxed_align, win_align);

...like in the other case.

-- 
 i.