All of lore.kernel.org
 help / color / mirror / Atom feed
From: Will Deacon <will@kernel.org>
To: Lu Baolu <baolu.lu@linux.intel.com>
Cc: Joerg Roedel <joro@8bytes.org>, Ashok Raj <ashok.raj@intel.com>,
	Jacob Pan <jacob.jun.pan@linux.intel.com>,
	Guo Kaijie <Kaijie.Guo@intel.com>, Liu Yi L <yi.l.liu@intel.com>,
	iommu@lists.linux-foundation.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH 3/5] iommu/vt-d: Remove unused dma map/unmap trace events
Date: Thu, 7 Jan 2021 14:40:50 +0000	[thread overview]
Message-ID: <20210107144049.GB3413@willie-the-truck> (raw)
In-Reply-To: <8bf33c93-3154-e832-1197-891c684d6ded@linux.intel.com>

On Wed, Jan 06, 2021 at 09:14:22AM +0800, Lu Baolu wrote:
> On 2021/1/6 3:04, Will Deacon wrote:
> > On Thu, Dec 31, 2020 at 08:53:21AM +0800, Lu Baolu wrote:
> > > With commit c588072bba6b5 ("iommu/vt-d: Convert intel iommu driver to
> > > the iommu ops"), the trace events for dma map/unmap have no users any
> > > more. Remove them so that they don't show up under
> > > /sys/kernel/debug/tracing/events/intel_iommu. The users should use the
> > > map/unmap traces defined in the iommu core from now on.
> > > 
> > > Fixes: c588072bba6b5 ("iommu/vt-d: Convert intel iommu driver to the iommu ops")
> > > Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
> > > ---
> > >   include/trace/events/intel_iommu.h | 119 -----------------------------
> > >   1 file changed, 119 deletions(-)
> > 
> > Is this needed in 5.11, or can it wait until 5.12?
> 
> It's necessary for 5.11 as far as I can see. Without this, users still
> see the events under /sys/kernel/debug/tracing/events/intel_iommu, but
> they will get nothing traced even they enable the events.

I'm just a bit wary about breaking userspace by removing them altogether,
although I see that there's plenty of precedent for removing events from
the include/trace/events directory, so it's probably fine.

However, the patch as-is results in this warning for me:

 | In file included from include/trace/define_trace.h:102,
 |                  from include/trace/events/intel_iommu.h:22,
 |                  from drivers/iommu/intel/trace.c:14:
 | include/trace/trace_events.h:27:23: warning: ‘str__intel_iommu__trace_system_name’ defined but not used [-Wunused-const-variable=]
 |    27 | #define __app__(x, y) str__##x##y
 |       |                       ^~~~~
 | include/trace/trace_events.h:28:21: note: in expansion of macro ‘__app__’
 |    28 | #define __app(x, y) __app__(x, y)
 |       |                     ^~~~~~~
 | include/trace/trace_events.h:30:29: note: in expansion of macro ‘__app’
 |    30 | #define TRACE_SYSTEM_STRING __app(TRACE_SYSTEM_VAR,__trace_system_name)
 |       |                             ^~~~~
 | include/trace/trace_events.h:33:20: note: in expansion of macro ‘TRACE_SYSTEM_STRING’
 |    33 |  static const char TRACE_SYSTEM_STRING[] = \
 |       |                    ^~~~~~~~~~~~~~~~~~~
 | include/trace/trace_events.h:36:1: note: in expansion of macro ‘TRACE_MAKE_SYSTEM_STR’
 |    36 | TRACE_MAKE_SYSTEM_STR();
 |       | ^~~~~~~~~~~~~~~~~~~~~

so I'll drop this for now.

Will

WARNING: multiple messages have this Message-ID (diff)
From: Will Deacon <will@kernel.org>
To: Lu Baolu <baolu.lu@linux.intel.com>
Cc: Ashok Raj <ashok.raj@intel.com>,
	linux-kernel@vger.kernel.org, iommu@lists.linux-foundation.org,
	Guo Kaijie <Kaijie.Guo@intel.com>
Subject: Re: [PATCH 3/5] iommu/vt-d: Remove unused dma map/unmap trace events
Date: Thu, 7 Jan 2021 14:40:50 +0000	[thread overview]
Message-ID: <20210107144049.GB3413@willie-the-truck> (raw)
In-Reply-To: <8bf33c93-3154-e832-1197-891c684d6ded@linux.intel.com>

On Wed, Jan 06, 2021 at 09:14:22AM +0800, Lu Baolu wrote:
> On 2021/1/6 3:04, Will Deacon wrote:
> > On Thu, Dec 31, 2020 at 08:53:21AM +0800, Lu Baolu wrote:
> > > With commit c588072bba6b5 ("iommu/vt-d: Convert intel iommu driver to
> > > the iommu ops"), the trace events for dma map/unmap have no users any
> > > more. Remove them so that they don't show up under
> > > /sys/kernel/debug/tracing/events/intel_iommu. The users should use the
> > > map/unmap traces defined in the iommu core from now on.
> > > 
> > > Fixes: c588072bba6b5 ("iommu/vt-d: Convert intel iommu driver to the iommu ops")
> > > Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
> > > ---
> > >   include/trace/events/intel_iommu.h | 119 -----------------------------
> > >   1 file changed, 119 deletions(-)
> > 
> > Is this needed in 5.11, or can it wait until 5.12?
> 
> It's necessary for 5.11 as far as I can see. Without this, users still
> see the events under /sys/kernel/debug/tracing/events/intel_iommu, but
> they will get nothing traced even they enable the events.

I'm just a bit wary about breaking userspace by removing them altogether,
although I see that there's plenty of precedent for removing events from
the include/trace/events directory, so it's probably fine.

However, the patch as-is results in this warning for me:

 | In file included from include/trace/define_trace.h:102,
 |                  from include/trace/events/intel_iommu.h:22,
 |                  from drivers/iommu/intel/trace.c:14:
 | include/trace/trace_events.h:27:23: warning: ‘str__intel_iommu__trace_system_name’ defined but not used [-Wunused-const-variable=]
 |    27 | #define __app__(x, y) str__##x##y
 |       |                       ^~~~~
 | include/trace/trace_events.h:28:21: note: in expansion of macro ‘__app__’
 |    28 | #define __app(x, y) __app__(x, y)
 |       |                     ^~~~~~~
 | include/trace/trace_events.h:30:29: note: in expansion of macro ‘__app’
 |    30 | #define TRACE_SYSTEM_STRING __app(TRACE_SYSTEM_VAR,__trace_system_name)
 |       |                             ^~~~~
 | include/trace/trace_events.h:33:20: note: in expansion of macro ‘TRACE_SYSTEM_STRING’
 |    33 |  static const char TRACE_SYSTEM_STRING[] = \
 |       |                    ^~~~~~~~~~~~~~~~~~~
 | include/trace/trace_events.h:36:1: note: in expansion of macro ‘TRACE_MAKE_SYSTEM_STR’
 |    36 | TRACE_MAKE_SYSTEM_STR();
 |       | ^~~~~~~~~~~~~~~~~~~~~

so I'll drop this for now.

Will
_______________________________________________
iommu mailing list
iommu@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/iommu

  reply	other threads:[~2021-01-07 14:41 UTC|newest]

Thread overview: 30+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-12-31  0:53 [PATCH 1/5] iommu/vt-d: Fix misuse of ALIGN in qi_flush_piotlb() Lu Baolu
2020-12-31  0:53 ` Lu Baolu
2020-12-31  0:53 ` [PATCH 2/5] iommu/vt-d: Fix unaligned addresses for intel_flush_svm_range_dev() Lu Baolu
2020-12-31  0:53   ` Lu Baolu
2021-01-05 19:03   ` Will Deacon
2021-01-05 19:03     ` Will Deacon
2021-01-06  1:09     ` Lu Baolu
2021-01-06  1:09       ` Lu Baolu
2021-01-07 23:52       ` Lu Baolu
2021-01-07 23:52         ` Lu Baolu
2021-01-08 14:09         ` Will Deacon
2021-01-08 14:09           ` Will Deacon
2021-01-08 14:30           ` Lu Baolu
2021-01-08 14:30             ` Lu Baolu
2020-12-31  0:53 ` [PATCH 3/5] iommu/vt-d: Remove unused dma map/unmap trace events Lu Baolu
2020-12-31  0:53   ` Lu Baolu
2021-01-05 19:04   ` Will Deacon
2021-01-05 19:04     ` Will Deacon
2021-01-06  1:14     ` Lu Baolu
2021-01-06  1:14       ` Lu Baolu
2021-01-07 14:40       ` Will Deacon [this message]
2021-01-07 14:40         ` Will Deacon
2021-01-08  0:00         ` Lu Baolu
2021-01-08  0:00           ` Lu Baolu
2020-12-31  0:53 ` [PATCH 4/5] Revert "iommu: Add quirk for Intel graphic devices in map_sg" Lu Baolu
2020-12-31  0:53   ` Lu Baolu
2020-12-31  0:53 ` [PATCH 5/5] iommu/vt-d: Fix lockdep splat in sva bind()/unbind() Lu Baolu
2020-12-31  0:53   ` Lu Baolu
2021-01-07 14:22 ` [PATCH 1/5] iommu/vt-d: Fix misuse of ALIGN in qi_flush_piotlb() Will Deacon
2021-01-07 14:22   ` Will Deacon

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=20210107144049.GB3413@willie-the-truck \
    --to=will@kernel.org \
    --cc=Kaijie.Guo@intel.com \
    --cc=ashok.raj@intel.com \
    --cc=baolu.lu@linux.intel.com \
    --cc=iommu@lists.linux-foundation.org \
    --cc=jacob.jun.pan@linux.intel.com \
    --cc=joro@8bytes.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=yi.l.liu@intel.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 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.