linux-kernel - Re: [PATCH] dma-remap: fix dma_common_find

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

Message-ID: <20251215204956.20659-1-andrei.popa105@yahoo.com>
Date: Mon, 15 Dec 2025 22:49:55 +0200
From: Andrei-Edward Popa <andrei.popa105@...oo.com>
To: robin.murphy@....com,
	m.szyprowski@...sung.com,
	hch@...radead.org
Cc: andrei.popa105@...oo.com,
	iommu@...ts.linux.dev,
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH] dma-remap: fix dma_common_find_pages() page lookup for offsets

Hi Christoph, Marek, Robin,

Thanks for the detailed explanations, that clarified things a lot.

I now understand that dma_mmap_attrs() expects cpu_addr to be the exact
pointer returned by dma_alloc_attrs(), and that subdividing an allocation
(e.g. via gen_pool) requires the driver to keep track of offsets and
adjust the mmap parameters accordingly before calling into the DMA API.

My patch was effectively trying to make dma_common_find_pages() accept
offset cpu_addr values implicitly, which would change the DMA API
semantics and mask driver bugs rather than fixing them in the driver
itself.

Given this, I agree that the correct fix is in the driver using the API,
not in dma-remap, and I’ll drop this patch.

Thanks again for taking the time to explain the intended usage and
constraints of the DMA mmap helpers.

Best regards,
Andrei