linux-kernel - swiotlb=force fix

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>] [day] [month] [year] [list]

Date:	Thu, 27 Jul 2006 13:15:34 +0300
From:	Muli Ben-Yehuda <muli@...ibm.com>
To:	Andi Kleen <ak@...e.de>
Cc:	Jon Mason <jdmason@...ibm.com>, linux-kernel@...r.kernel.org,
	discuss@...-64.org
Subject: swiotlb=force fix

Hi Andi,

I was looking at your swiotlb=force fix at
ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.18-rc2/2.6.18-rc2-mm1/broken-out/x86_64-mm-fix-swiotlb-force.patch.

In general I agree with the patch (and thanks for fixing it).

> -	if (!iommu_detected && !no_iommu &&
> -	    (end_pfn > MAX_DMA32_PFN || force_iommu))
> +	if (!iommu_detected && !no_iommu && end_pfn > MAX_DMA32_PFN)
>  	       swiotlb = 1;

what is the expected outcome if the user specifies 'iommu=force' with
no HW IOMMU in the machine and swiotlb compiled in? we thought the
path of least surprise is to enable swiotlb unconditionally in this
case. You removed the check for force_iommu here so now this set of
command line options will *not* turn on swiotlb unless there's more
than 4G in the machine, which can be surprising.

> +extern int swiotlb_force;
> +
>  #ifdef CONFIG_SWIOTLB
>  extern int swiotlb;
>  #else
>  #define swiotlb 0
>  #endif

Any reason not to move this into the CONFIG_SWIOTLB bit the same way
'extern int swiotlb' is declared?

Cheers,
Muli
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/