From: Alexey Kardashevskiy <aik@ozlabs.ru>
To: Leonardo Bras <leobras.c@gmail.com>,
Michael Ellerman <mpe@ellerman.id.au>,
Benjamin Herrenschmidt <benh@kernel.crashing.org>,
Paul Mackerras <paulus@samba.org>,
Christophe Leroy <christophe.leroy@c-s.fr>,
Joel Stanley <joel@jms.id.au>,
Thiago Jung Bauermann <bauerman@linux.ibm.com>,
Ram Pai <linuxram@us.ibm.com>,
Brian King <brking@linux.vnet.ibm.com>,
Murilo Fossa Vicentini <muvic@linux.ibm.com>,
David Dai <zdai@linux.vnet.ibm.com>
Cc: linuxppc-dev@lists.ozlabs.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH v1 02/10] powerpc/kernel/iommu: Align size for IOMMU_PAGE_SIZE on iommu_*_coherent()
Date: Tue, 8 Sep 2020 13:18:44 +1000 [thread overview]
Message-ID: <bafc94fe-85e8-88be-a75d-45293b56fa09@ozlabs.ru> (raw)
In-Reply-To: <ef7e80b0a7399bad607324301a604bfb46c2de05.camel@gmail.com>
On 04/09/2020 16:04, Leonardo Bras wrote:
> On Thu, 2020-09-03 at 14:41 +1000, Alexey Kardashevskiy wrote:
>> I am new to this, so I am trying to understand how a memory page mapped
>>> as DMA, and used for something else could be a problem.
>>
>> From the device prospective, there is PCI space and everything from 0
>> till 1<<64 is accessible and what is that mapped to - the device does
>> not know. PHB's IOMMU is the thing to notice invalid access and raise
>> EEH but PHB only knows about PCI->physical memory mapping (with IOMMU
>> pages) but nothing about the host kernel pages. Does this help? Thanks,
>
> According to our conversation on Slack:
> 1- There is a problem if a hypervisor gives to it's VMs contiguous
> memory blocks that are not aligned to IOMMU pages, because then an
> iommu_map_page() could map some memory in this VM and some memory in
> other VM / process.
> 2- To guarantee this, we should have system pagesize >= iommu_pagesize
>
> One way to get (2) is by doing this in enable_ddw():
> if ((query.page_size & 4) && PAGE_SHIFT >= 24) {
You won't ever (well, soon) see PAGE_SHIFT==24, it is either 4K or 64K.
However 16MB IOMMU pages is fine - if hypervisor uses huge pages for VMs
RAM, it also then advertises huge IOMMU pages in ddw-query. So for the
1:1 case there must be no "PAGE_SHIFT >= 24".
> page_shift = 24; /* 16MB */
> } else if ((query.page_size & 2) && PAGE_SHIFT >= 16 ) {
> page_shift = 16; /* 64kB */
> } else if (query.page_size & 1 && PAGE_SHIFT >= 12) {
> page_shift = 12; /* 4kB */
> [...]
>
> Another way of solving this, would be adding in LoPAR documentation
> that the blocksize of contiguous memory the hypervisor gives a VM
> should always be aligned to IOMMU pagesize offered.
I think this is assumed already by the design of the DDW API.
>
> I think the best approach would be first sending the above patch, which
> is faster, and then get working into adding that to documentation, so
> hypervisors guarantee this.
>
> If this gets into the docs, we can revert the patch.
>
> What do you think?
I think we diverted from the original patch :) I am not quite sure what
you were fixing there. Thanks,
--
Alexey
next prev parent reply other threads:[~2020-09-08 3:20 UTC|newest]
Thread overview: 52+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-08-17 23:40 [PATCH v1 00/10] DDW indirect mapping Leonardo Bras
2020-08-17 23:40 ` [PATCH v1 01/10] powerpc/pseries/iommu: Replace hard-coded page shift Leonardo Bras
2020-08-22 9:33 ` Alexey Kardashevskiy
2020-08-27 15:32 ` Leonardo Bras
2020-08-28 2:27 ` Alexey Kardashevskiy
2020-08-28 19:55 ` Leonardo Bras
2020-08-31 0:06 ` Alexey Kardashevskiy
2020-08-31 1:41 ` Oliver O'Halloran
2020-08-31 3:48 ` Alexey Kardashevskiy
2020-09-01 21:38 ` Leonardo Bras
2020-09-03 4:26 ` Alexey Kardashevskiy
2020-08-17 23:40 ` [PATCH v1 02/10] powerpc/kernel/iommu: Align size for IOMMU_PAGE_SIZE on iommu_*_coherent() Leonardo Bras
2020-08-22 10:07 ` Alexey Kardashevskiy
2020-08-27 16:51 ` Leonardo Bras
2020-08-28 1:40 ` Alexey Kardashevskiy
2020-08-28 20:41 ` Leonardo Bras
2020-08-31 0:47 ` Alexey Kardashevskiy
2020-09-01 22:34 ` Leonardo Bras
2020-09-03 4:41 ` Alexey Kardashevskiy
2020-09-04 6:04 ` Leonardo Bras
2020-09-08 3:18 ` Alexey Kardashevskiy [this message]
2020-08-17 23:40 ` [PATCH v1 03/10] powerpc/kernel/iommu: Use largepool as a last resort when !largealloc Leonardo Bras
2020-08-22 10:09 ` Alexey Kardashevskiy
2020-08-27 16:58 ` Leonardo Bras
2020-08-17 23:40 ` [PATCH v1 04/10] powerpc/kernel/iommu: Add new iommu_table_in_use() helper Leonardo Bras
2020-08-22 10:34 ` Alexey Kardashevskiy
2020-08-27 18:34 ` Leonardo Bras
2020-08-28 1:51 ` Alexey Kardashevskiy
2020-08-17 23:40 ` [PATCH v1 05/10] powerpc/pseries/iommu: Add iommu_pseries_alloc_table() helper Leonardo Bras
2020-08-24 0:38 ` Alexey Kardashevskiy
2020-08-27 21:23 ` Leonardo Bras
2020-08-17 23:40 ` [PATCH v1 06/10] powerpc/pseries/iommu: Add ddw_list_add() helper Leonardo Bras
2020-08-24 3:46 ` Alexey Kardashevskiy
2020-08-27 22:11 ` Leonardo Bras
2020-08-28 1:58 ` Alexey Kardashevskiy
2020-08-28 21:28 ` Leonardo Bras
2020-08-17 23:40 ` [PATCH v1 07/10] powerpc/pseries/iommu: Allow DDW windows starting at 0x00 Leonardo Bras
2020-08-24 3:44 ` Alexey Kardashevskiy
2020-08-28 14:04 ` Leonardo Bras
2020-08-31 0:50 ` Alexey Kardashevskiy
2020-08-17 23:40 ` [PATCH v1 08/10] powerpc/pseries/iommu: Add ddw_property_create() and refactor enable_ddw() Leonardo Bras
2020-08-24 5:07 ` Alexey Kardashevskiy
2020-08-28 15:25 ` Leonardo Bras
2020-08-31 4:34 ` Alexey Kardashevskiy
2020-09-02 5:27 ` Leonardo Bras
2020-08-17 23:40 ` [PATCH v1 09/10] powerpc/pseries/iommu: Make use of DDW even if it does not map the partition Leonardo Bras
2020-08-24 5:17 ` Alexey Kardashevskiy
2020-08-28 18:36 ` Leonardo Bras
2020-08-31 4:35 ` Alexey Kardashevskiy
2020-09-02 6:11 ` Leonardo Bras
2020-09-04 1:00 ` Alexey Kardashevskiy
2020-08-17 23:40 ` [PATCH v1 10/10] powerpc/pseries/iommu: Rename "direct window" to "dma window" Leonardo Bras
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=bafc94fe-85e8-88be-a75d-45293b56fa09@ozlabs.ru \
--to=aik@ozlabs.ru \
--cc=bauerman@linux.ibm.com \
--cc=benh@kernel.crashing.org \
--cc=brking@linux.vnet.ibm.com \
--cc=christophe.leroy@c-s.fr \
--cc=joel@jms.id.au \
--cc=leobras.c@gmail.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=linuxram@us.ibm.com \
--cc=mpe@ellerman.id.au \
--cc=muvic@linux.ibm.com \
--cc=paulus@samba.org \
--cc=zdai@linux.vnet.ibm.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).