From: Suravee Suthikulpanit <Suravee.Suthikulpanit-5C7GfCeVMHo@public.gmane.org>
To: Alex Williamson
<alex.williamson-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
Cc: Peter Zijlstra <peterz-wEGCiKHe2LqWVfeAwA7xHQ@public.gmane.org>,
iommu-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org,
acme-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org,
linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
andihartmann-KuiJ5kEpwI6ELgA04lAiVw@public.gmane.org,
mingo-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org,
bp-Gina5bIWoIWzQB+pC5nmwQ@public.gmane.org
Subject: Re: [PATCH V4 5/6] perf/amd/iommu: Enable support for multiple IOMMUs
Date: Tue, 23 Feb 2016 16:56:22 +0700 [thread overview]
Message-ID: <56CC2CC6.6090803@amd.com> (raw)
In-Reply-To: <20160222222457.2824f7e0-1yVPhWWZRC1BDLzU/O5InQ@public.gmane.org>
On 02/23/2016 12:24 PM, Alex Williamson wrote:
> On Tue, 23 Feb 2016 12:12:42 +0700
> Suravee Suthikulpanit <Suravee.Suthikulpanit-5C7GfCeVMHo@public.gmane.org> wrote:
>
>> Hi
>>
>> On 02/22/2016 09:07 PM, Peter Zijlstra wrote:
>>> On Mon, Feb 22, 2016 at 03:00:31PM +0700, Suravee Suthikulpanit wrote:
>>>>> So I really don't have time to review new muck while I'm hunting perf
>>>>> core fail, but Boris made me look at this.
>>>>>
>>>>> This is crazy, if you have multiple IOMMUs then create an event per
>>>>> IOMMU, do _NOT_ fold them all into a single event.
>>>>
>>>> These are system-wide events, which are programmed on every IOMMU the same
>>>> way. I am not sure what you meant by creating an event per IOMMU. Do you
>>>> mean I should create internal per-IOMMU struct perf_event for each event?
>>>
>>> No, I meant to expose each IOMMU individually to userspace, as a
>>> separate device.
>>>
>>> Is there never a case to profile just one of the IOMMUs ?
>>>
>>
>> I see. That's definitely doable and simpler to implement.
>>
>> I was not sure if making users specify the IOMMU instance (e.g.
>> amd_iommu_0/<ev name> , amd_iommu_1/<ev_name>, ....) would be too
>> tedious. However, this would actually give users better control of the
>> performance events, which is a good trade-off. I think it is acceptable.
>>
>> I'll make the change and send this out in V5.
>
> We already expose individual IOMMU hardware units in /sys/class/iommu/,
> you might consider trying to match the names there for the convenience
> of the user. Looks like we use ivhd%d for AMD. Thanks,
>
> Alex
>
Hm, the PMU for AMD IOMMU is currently shown as /sys/device/amd_iommu. I
am not sure if /sys/device/ivhd[0|1|...] would be obvious to users that
IVHD is really the AMD IOMMU.
Suravee
WARNING: multiple messages have this Message-ID (diff)
From: Suravee Suthikulpanit <Suravee.Suthikulpanit@amd.com>
To: Alex Williamson <alex.williamson@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>,
<iommu@lists.linux-foundation.org>,
<linux-kernel@vger.kernel.org>, <acme@kernel.org>,
<andihartmann@freenet.de>, <mingo@redhat.com>, <bp@alien8.de>
Subject: Re: [PATCH V4 5/6] perf/amd/iommu: Enable support for multiple IOMMUs
Date: Tue, 23 Feb 2016 16:56:22 +0700 [thread overview]
Message-ID: <56CC2CC6.6090803@amd.com> (raw)
In-Reply-To: <20160222222457.2824f7e0@t450s.home>
On 02/23/2016 12:24 PM, Alex Williamson wrote:
> On Tue, 23 Feb 2016 12:12:42 +0700
> Suravee Suthikulpanit <Suravee.Suthikulpanit@amd.com> wrote:
>
>> Hi
>>
>> On 02/22/2016 09:07 PM, Peter Zijlstra wrote:
>>> On Mon, Feb 22, 2016 at 03:00:31PM +0700, Suravee Suthikulpanit wrote:
>>>>> So I really don't have time to review new muck while I'm hunting perf
>>>>> core fail, but Boris made me look at this.
>>>>>
>>>>> This is crazy, if you have multiple IOMMUs then create an event per
>>>>> IOMMU, do _NOT_ fold them all into a single event.
>>>>
>>>> These are system-wide events, which are programmed on every IOMMU the same
>>>> way. I am not sure what you meant by creating an event per IOMMU. Do you
>>>> mean I should create internal per-IOMMU struct perf_event for each event?
>>>
>>> No, I meant to expose each IOMMU individually to userspace, as a
>>> separate device.
>>>
>>> Is there never a case to profile just one of the IOMMUs ?
>>>
>>
>> I see. That's definitely doable and simpler to implement.
>>
>> I was not sure if making users specify the IOMMU instance (e.g.
>> amd_iommu_0/<ev name> , amd_iommu_1/<ev_name>, ....) would be too
>> tedious. However, this would actually give users better control of the
>> performance events, which is a good trade-off. I think it is acceptable.
>>
>> I'll make the change and send this out in V5.
>
> We already expose individual IOMMU hardware units in /sys/class/iommu/,
> you might consider trying to match the names there for the convenience
> of the user. Looks like we use ivhd%d for AMD. Thanks,
>
> Alex
>
Hm, the PMU for AMD IOMMU is currently shown as /sys/device/amd_iommu. I
am not sure if /sys/device/ivhd[0|1|...] would be obvious to users that
IVHD is really the AMD IOMMU.
Suravee
next prev parent reply other threads:[~2016-02-23 9:56 UTC|newest]
Thread overview: 43+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-02-11 9:15 [PATCH V4 0/6] perf/amd/iommu: Enable multi-IOMMU support Suravee Suthikulpanit
2016-02-11 9:15 ` Suravee Suthikulpanit
2016-02-11 9:15 ` [PATCH V4 3/6] iommu/amd: Introduce amd_iommu_get_num_iommus() Suravee Suthikulpanit
2016-02-11 9:15 ` Suravee Suthikulpanit
[not found] ` <1455182127-17551-1-git-send-email-Suravee.Suthikulpanit-5C7GfCeVMHo@public.gmane.org>
2016-02-11 9:15 ` [PATCH V4 1/6] perf/amd/iommu: Consolidate and move perf_event_amd_iommu header Suravee Suthikulpanit
2016-02-11 9:15 ` Suravee Suthikulpanit
2016-02-11 9:15 ` [PATCH V4 2/6] perf/amd/iommu: Modify functions to query max banks and counters Suravee Suthikulpanit
2016-02-11 9:15 ` Suravee Suthikulpanit
2016-02-18 11:11 ` Borislav Petkov
[not found] ` <20160218111116.GE3694-fF5Pk5pvG8Y@public.gmane.org>
2016-02-22 4:55 ` Suravee Suthikulpanit
2016-02-22 4:55 ` Suravee Suthikulpanit
2016-02-11 9:15 ` [PATCH V4 4/6] perf/amd/iommu: Introduce get_iommu_bnk_cnt_evt_idx Suravee Suthikulpanit
2016-02-11 9:15 ` Suravee Suthikulpanit
[not found] ` <1455182127-17551-5-git-send-email-Suravee.Suthikulpanit-5C7GfCeVMHo@public.gmane.org>
2016-02-18 11:45 ` Borislav Petkov
2016-02-18 11:45 ` Borislav Petkov
2016-02-11 9:15 ` [PATCH V4 5/6] perf/amd/iommu: Enable support for multiple IOMMUs Suravee Suthikulpanit
2016-02-11 9:15 ` Suravee Suthikulpanit
2016-02-18 13:18 ` Peter Zijlstra
[not found] ` <20160218131853.GU6357-ndre7Fmf5hadTX5a5knrm8zTDFooKrT+cvkQGrU6aU0@public.gmane.org>
2016-02-22 8:00 ` Suravee Suthikulpanit
2016-02-22 8:00 ` Suravee Suthikulpanit
2016-02-22 14:07 ` Peter Zijlstra
[not found] ` <20160222140741.GH6357-ndre7Fmf5hadTX5a5knrm8zTDFooKrT+cvkQGrU6aU0@public.gmane.org>
2016-02-23 5:12 ` Suravee Suthikulpanit
2016-02-23 5:12 ` Suravee Suthikulpanit
[not found] ` <56CBEA4A.8070001-5C7GfCeVMHo@public.gmane.org>
2016-02-23 5:24 ` Alex Williamson
2016-02-23 5:24 ` Alex Williamson
[not found] ` <20160222222457.2824f7e0-1yVPhWWZRC1BDLzU/O5InQ@public.gmane.org>
2016-02-23 9:56 ` Suravee Suthikulpanit [this message]
2016-02-23 9:56 ` Suravee Suthikulpanit
[not found] ` <1455182127-17551-6-git-send-email-Suravee.Suthikulpanit-5C7GfCeVMHo@public.gmane.org>
2016-02-18 13:21 ` Borislav Petkov
2016-02-18 13:21 ` Borislav Petkov
2016-02-11 9:15 ` [PATCH V4 6/6] perf/amd/iommu: Clean up print messages pr_debug Suravee Suthikulpanit
2016-02-11 9:15 ` Suravee Suthikulpanit
[not found] ` <1455182127-17551-7-git-send-email-Suravee.Suthikulpanit-5C7GfCeVMHo@public.gmane.org>
2016-02-18 13:22 ` Borislav Petkov
2016-02-18 13:22 ` Borislav Petkov
2016-02-18 2:30 ` [PATCH V4 0/6] perf/amd/iommu: Enable multi-IOMMU support Suravee Suthikulpanit
2016-02-18 2:30 ` Suravee Suthikulpanit
2016-02-23 11:04 ` Joerg Roedel
2016-02-23 11:04 ` Joerg Roedel
[not found] ` <20160223110406.GA22747-zLv9SwRftAIdnm+yROfE0A@public.gmane.org>
2016-02-23 11:27 ` Suravee Suthikulpanit
2016-02-23 11:27 ` Suravee Suthikulpanit
[not found] ` <56CC422E.7050905-5C7GfCeVMHo@public.gmane.org>
2016-02-23 11:39 ` Suravee Suthikulpanit
2016-02-23 11:39 ` Suravee Suthikulpanit
[not found] ` <56CC44D8.4080604-5C7GfCeVMHo@public.gmane.org>
2016-02-23 12:12 ` [PATCH] iommu/amd: Fix boot warning when device 00:00.0 is not iommu Joerg Roedel
2016-02-23 12:12 ` Joerg Roedel
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=56CC2CC6.6090803@amd.com \
--to=suravee.suthikulpanit-5c7gfcevmho@public.gmane.org \
--cc=acme-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org \
--cc=alex.williamson-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org \
--cc=andihartmann-KuiJ5kEpwI6ELgA04lAiVw@public.gmane.org \
--cc=bp-Gina5bIWoIWzQB+pC5nmwQ@public.gmane.org \
--cc=iommu-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org \
--cc=linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
--cc=mingo-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org \
--cc=peterz-wEGCiKHe2LqWVfeAwA7xHQ@public.gmane.org \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.