* Re: [PATCH 07/18] media controller: rename the tuner entity
From: Hans Verkuil @ 2015-05-08 14:32 UTC (permalink / raw)
To: Mauro Carvalho Chehab
Cc: Linux Media Mailing List, Mauro Carvalho Chehab, Jonathan Corbet,
Matthias Schwarzott, Antti Palosaari, Olli Salonen, Prabhakar Lad,
Sakari Ailus, Laurent Pinchart, linux-doc-u79uwXL29TY76Z2rM5mHXA,
linux-api-u79uwXL29TY76Z2rM5mHXA
In-Reply-To: <20150508110826.00e4e954-+RedX5hVuTR+urZeOPWqwQ@public.gmane.org>
On 05/08/2015 04:08 PM, Mauro Carvalho Chehab wrote:
> Em Fri, 08 May 2015 15:21:39 +0200
> Hans Verkuil <hverkuil-qWit8jRvyhVmR6Xm/wNWPw@public.gmane.org> escreveu:
>
>> On 05/08/2015 02:57 PM, Mauro Carvalho Chehab wrote:
>>> Em Fri, 08 May 2015 14:13:22 +0200
>>> Hans Verkuil <hverkuil-qWit8jRvyhVmR6Xm/wNWPw@public.gmane.org> escreveu:
>>>
>>>> On 05/08/2015 03:12 AM, Mauro Carvalho Chehab wrote:
>>>>> Finally, let's rename the tuner entity. inside the media subsystem,
>>>>> a tuner can be used by AM/FM radio, SDR radio, analog TV and digital TV.
>>>>> It could even be used on other subsystems, like network, for wireless
>>>>> devices.
>>>>>
>>>>> So, it is not constricted to V4L2 API, or to a subdev.
>>>>>
>>>>> Let's then rename it as:
>>>>> MEDIA_ENT_T_V4L2_SUBDEV_TUNER -> MEDIA_ENT_T_TUNER
>>>>
>>>> See patch 04/18.
>>>
>>> Mapping the tuner as a V4L2_SUBDEV is plain wrong. We can't assume
>>> that a tuner will always be mapped via V4L2 subdev API.
>>
>> True. Today we have subdevs that have no device node to control them, so
>> in that case it would just be a SUBDEV entity. There are subdevs that make
>> a v4l-subdev device node, so those can be V4L(2)_SUBDEV entities.
>>
>> The question is: what are your ideas for e.g. DVB-only tuners? Would they
>> get a DVB-like device node? (so DTV_SUBDEV)
>
> I guess we may need DVB subdevs in the future. For now, I don't see
> much usage.
>
>> Would hybrid tuners have two
>> device nodes? One v4l-subdev, one dvb/dtv-subdev?
>
> No. A tuner is a tuner. The very same device can be used for analog or
> digital TV. Ok, there are tuners that only work for digital TV (satellite
> tuners, typically), because satellite requires a different tuning range,
> and require an extra hardware to power up the satellite antena. So, on
> most devices, the tuner is integrated with SEC.
>
> In any case, I don't see any reason why artificially one piece of hardware
> component (tuner) into one subdevice entity per API.
>
> What it might make sense in the future is to have some DVB-specific ioctls
> for a hybrid tuner, in order to allow adjusting its internal filters to
> an specific digital TV standard.
>
>> Just curious what your thoughts are.
>>
>> Brainstorming:
>>
>> It might be better to map each device node to an entity and each hardware
>> component (tuner, DMA engine) to an entity, and avoid this mixing of
>> hw entity vs device node entity.
>
> Ok, but then we need to properly define the namespaces for HW and for
> Linux API components.
>
> So, we would have a namespace like:
>
> - ENT_T_DEVNODE_DVB_(FE|CA|NET|DEMUX|DVR) for the DVB API device nodes;
> - ENT_T_DEVNODE_V4L for the radio/swradio/video/vbi devnodes, or,
> alternatively, use ENT_T_DEVNODE_V4L_(RADIO|SWRADIO|VIDEO|VBI);
> - ENT_T_HW_(TUNER|CAM_SENSOR|ATV_DEMOD|DTV_DEMOD|...) for the
> hardware components.
>
> In other words, the namespace would actually define two subtypes:
> - devnodes;
> - hardware components
>
> There's one advantage on this strategy: it is easier to keep backward
> compatibility, IMHO, as we'll be preserving 1 << 16 for device nodes
Right, that will work.
> and 2 << 16 for hardware components.
This remains problematic since I believe this should be done as a list
of properties. Instead the entity type would be ENT_T_HW (if it represents
actual hardware), ENT_T_SW (if it is a software implementation, for example
for the DVB demux if there is no HW demux), and perhaps ENT_T_IP for
representing IP blocks (not sure about this one).
And the properties will tell what functions it supports.
The existing 2 << 16 defines would only be used if the property list matches
the original meaning of the define to keep it backwards compatible.
>
> Yet, we'll need to add an entity for the V4L2 hardware DMA (with makes
> sense to me), and this might break backward compatibility if not done
> well.
I see this as a property as well, but otherwise I agree with this.
>
> It should be said that, in such case, hardware components will then
> mean not only V4L2-specific hardware (V4L2_SUBDEV_foo), but also DVB,
> ALSA, ... components.
>
> So, we'll still need a way to identify what of those components are
> V4L2 subdevs, probably using the properties API.
Why? A hardware component that can be controlled via a v4l-subdev node
would be linked to an entity for that v4l-subdev node. That's an API
entity. The whole 'is this a v4l-subdev' question disappears, since that
is now no longer relevant. Instead you will have an ENT_T_DEVNODE_V4L_SUBDEV
entity linked to the hw entity.
Even a radio device would fit cleanly into this: the tuner entity simply
has a link to a radio device node entity.
Hmm, does this also solve the control vs DMA issue? If a DEVNODE entity
is hooked up to an entity with the DMA functionality, then you can stream,
otherwise it is just for control.
I'm not sure if this is always true, though. Of course, we can also just
add the streaming/dma property to the DEVNODE entity as well.
> If you all agree with that, I'll respin the patch series to map the
> entities like that.
We can interpret the existing ENT_T_HW_TUNER etc. as shorthand for a
property while the property API isn't there yet (we need that anyway
for backwards compat).
So we would need to add a ENT_T_HW_DMA as well (to be replaced by a
property later).
Basically I see the 1 << 16 range as device node types, the 2 << 16 range
as shorthands for what should be properties (this really defined functions
and entities can combine multiple functions), and we would need to have a
new range (4 << 16) for non-DEVNODE entity types. Although we could keep
it in range 1 << 16 as well, but I think it might make sense to keep it
separate.
And there you would get ENT_T_HW/SW/IP (not sure about the last one). And
perhaps FPGA.
Again, just brainstorming here.
Regards,
Hans
^ permalink raw reply
* Re: [PATCH RFC] vfs: add a O_NOMTIME flag
From: Austin S Hemmelgarn @ 2015-05-08 14:43 UTC (permalink / raw)
To: Sage Weil, Zach Brown
Cc: Dave Chinner, Alexander Viro,
linux-fsdevel-u79uwXL29TY76Z2rM5mHXA,
linux-kernel-u79uwXL29TY76Z2rM5mHXA,
linux-api-u79uwXL29TY76Z2rM5mHXA
In-Reply-To: <alpine.DEB.2.00.1505071752520.28239-vIokxiIdD2AQNTJnQDzGJqxOck334EZe@public.gmane.org>
[-- Attachment #1: Type: text/plain, Size: 3051 bytes --]
On 2015-05-07 21:01, Sage Weil wrote:
> On Thu, 7 May 2015, Zach Brown wrote:
>> On Thu, May 07, 2015 at 10:26:17AM +1000, Dave Chinner wrote:
>>> On Wed, May 06, 2015 at 03:00:12PM -0700, Zach Brown wrote:
>>>> The criteria for using O_NOMTIME is the same as for using O_NOATIME:
>>>> owning the file or having the CAP_FOWNER capability. If we're not
>>>> comfortable allowing owners to prevent mtime/ctime updates then we
>>>> should add a tunable to allow O_NOMTIME. Maybe a mount option?
>>>
>>> I dislike "turn off safety for performance" options because Joe
>>> SpeedRacer will always select performance over safety.
>>
>> Well, for ceph there's no safety concern. They never use cmtime in
>> these files.
>>
>> So are you suggesting not implementing this and making them rework their
>> IO paths to avoid the fs maintaining mtime so that we don't give Joe
>> Speedracer more rope? Or are we talking about adding some speed bumps
>> that ceph can flip on that might give Joe Speedracer pause?
>
> I think this is the fundamental question: who do we give the ammunition
> to, the user or app writer, or the sysadmin?
>
> One might argue that we gave the user a similar power with O_NOATIME (the
> power to break applications that assume atime is accurate). Here we give
> developers/users the power to not update mtime and suffer the consequences
> (like, obviously, breaking mtime-based backups). It should be pretty
> obvious to anyone using the flag what the consequences are.
The difference is that the only widely used program that uses atime for
anything is Mutt (and many people who don't use Mutt just disable
updating it altogether to improve performance), whereas mtime is used at
the very least by many backup tools, and pretty much all NFSv{3,2}
clients, as well as a number of other pieces of software.
>
> Note that we can suffer similar lapses in mtime with fdatasync followed by
> a system crash. And as Andy points out it's semi-broken for writable
> mmap. The crash case is obviously a slightly different thing, but the
> idea that mtime can't always be trusted certainly isn't crazy talk.
>
> Or, we can be conservative and require a mount option so that the admin
> has to explicitly allow behavior that might break some existing
> assumptions about mtime/ctime ('-o user_noatime' I guess?).
Personally, I agree that there should be a mount option. We should make
sure to put a big fat warning about it in the manpage however,
irrespective of how it is controlled.
>
> I'm happy either way, so long as in the end an unprivileged ceph daemon
> avoids the useless work. In our case we always own the entire mount/disk,
> so a mount option is just fine.
> Thanks!
> sage
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
>
[-- Attachment #2: S/MIME Cryptographic Signature --]
[-- Type: application/pkcs7-signature, Size: 2967 bytes --]
^ permalink raw reply
* Re: [PATCH RFC] vfs: add a O_NOMTIME flag
From: Eric Sandeen @ 2015-05-08 14:44 UTC (permalink / raw)
To: Andy Lutomirski, Dave Chinner
Cc: Al Viro, Sage Weil, Linux API, Linux FS Devel,
linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, Zach Brown
In-Reply-To: <CALCETrUksu5ZB4QBfC8DMwYO2OFjfPW2eWsTweZGN_gybzcsmw-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
On 5/7/15 10:24 PM, Andy Lutomirski wrote:
> On May 8, 2015 8:11 AM, "Dave Chinner" <david-FqsqvQoI3Ljby3iVrkZq2A@public.gmane.org> wrote:
>>
>> On Thu, May 07, 2015 at 10:20:53AM -0700, Zach Brown wrote:
>>> On Thu, May 07, 2015 at 10:26:17AM +1000, Dave Chinner wrote:
>>>> On Wed, May 06, 2015 at 03:00:12PM -0700, Zach Brown wrote:
>>>>> Add the O_NOMTIME flag which prevents mtime from being updated which can
>>>>> greatly reduce the IO overhead of writes to allocated and initialized
>>>>> regions of files.
>>>>
>>>> Hmmm. How do backup programs now work out if the file has changed
>>>> and hence needs copying again? ie. applications using this will
>>>> break other critical infrastructure in subtle ways.
>>>
>>> By using backup infrastructure that doesn't use cmtime. Like btrfs
>>> send/recv. Or application level backups that know how to do
>>> incrementals from metadata in giant database files, say, without
>>> walking, comparing, and copying the entire thing.
>>
>> "Use magical thing that doesn't exist"? Really?
>>
>> e.g. you can't do incremental backups with tools like xfsdump if
>> mtime is not being updated. The last thing an admin wants when
>> doing disaster recovery is to find out that the app started using
>> O_NOMTIME as a result of the upgrade they did 6 months ago. Hence
>> the last 6 months of production data isn't in the backups despite
>> the backup procedure having been extensively tested and verified
>> when it was first put in place.
>>
>>>>> The criteria for using O_NOMTIME is the same as for using O_NOATIME:
>>>>> owning the file or having the CAP_FOWNER capability. If we're not
>>>>> comfortable allowing owners to prevent mtime/ctime updates then we
>>>>> should add a tunable to allow O_NOMTIME. Maybe a mount option?
>>>>
>>>> I dislike "turn off safety for performance" options because Joe
>>>> SpeedRacer will always select performance over safety.
>>>
>>> Well, for ceph there's no safety concern. They never use cmtime in
>>> these files.
>>
>> Understood.
>>
>>> So are you suggesting not implementing this
>>
>> No.
>>
>>> Or are we talking about adding some speed bumps
>>> that ceph can flip on that might give Joe Speedracer pause?
>>
>> Yes, but not just Joe Speedracer - if it can be turned on silently
>> by apps then it's a great big landmine that most users and sysadmins
>> will not know about until it is too late.
>
> What about programs like tar that explicitly override mtime? No admin
> buy-in is required for that. Admittedly, that doesn't affect ctime,
> nor is it as likely to bite unexpectedly as a nomtime flag.
>
> I think it would be reasonably safe if a mount option had to be set to
> allow O_NOCMTIME or such.
I was going to suggest the same. Make infrastructure available for an app
to request O_NOMTIME, but a mount option must be set to allow it, so the
administrator doesn't get an unhappy surprise at backup-restore time.
(Not a big fan of more twiddly knobs, but that seems to put the control
in all the right places).
-Eric
^ permalink raw reply
* Re: [PATCH RFC] vfs: add a O_NOMTIME flag
From: Sage Weil @ 2015-05-08 15:19 UTC (permalink / raw)
To: Trond Myklebust
Cc: Zach Brown, Dave Chinner, Alexander Viro,
Linux FS-devel Mailing List, Linux Kernel Mailing List,
Linux API Mailing List
In-Reply-To: <CAHQdGtQjMHA8rVPkggB2zMz=k3O667+APH_1EY_2FtYmHL7-hw@mail.gmail.com>
On Thu, 7 May 2015, Trond Myklebust wrote:
> On Thu, May 7, 2015 at 9:01 PM, Sage Weil <sage@newdream.net> wrote:
> > On Thu, 7 May 2015, Zach Brown wrote:
> >> On Thu, May 07, 2015 at 10:26:17AM +1000, Dave Chinner wrote:
> >> > On Wed, May 06, 2015 at 03:00:12PM -0700, Zach Brown wrote:
> >> > > The criteria for using O_NOMTIME is the same as for using O_NOATIME:
> >> > > owning the file or having the CAP_FOWNER capability. If we're not
> >> > > comfortable allowing owners to prevent mtime/ctime updates then we
> >> > > should add a tunable to allow O_NOMTIME. Maybe a mount option?
> >> >
> >> > I dislike "turn off safety for performance" options because Joe
> >> > SpeedRacer will always select performance over safety.
> >>
> >> Well, for ceph there's no safety concern. They never use cmtime in
> >> these files.
> >>
> >> So are you suggesting not implementing this and making them rework their
> >> IO paths to avoid the fs maintaining mtime so that we don't give Joe
> >> Speedracer more rope? Or are we talking about adding some speed bumps
> >> that ceph can flip on that might give Joe Speedracer pause?
> >
> > I think this is the fundamental question: who do we give the ammunition
> > to, the user or app writer, or the sysadmin?
> >
> > One might argue that we gave the user a similar power with O_NOATIME (the
> > power to break applications that assume atime is accurate). Here we give
> > developers/users the power to not update mtime and suffer the consequences
> > (like, obviously, breaking mtime-based backups). It should be pretty
> > obvious to anyone using the flag what the consequences are.
> >
> > Note that we can suffer similar lapses in mtime with fdatasync followed by
> > a system crash. And as Andy points out it's semi-broken for writable
> > mmap. The crash case is obviously a slightly different thing, but the
> > idea that mtime can't always be trusted certainly isn't crazy talk.
> >
> > Or, we can be conservative and require a mount option so that the admin
> > has to explicitly allow behavior that might break some existing
> > assumptions about mtime/ctime ('-o user_noatime' I guess?).
> >
> > I'm happy either way, so long as in the end an unprivileged ceph daemon
> > avoids the useless work. In our case we always own the entire mount/disk,
> > so a mount option is just fine.
> >
>
> So, what is the expectation here for filesystems that cannot support
> this flag? NFSv3 in particular would break pretty catastrophically if
> someone decided on a whim to turn off mtime: they will have turned off
> the client's ability to detect cache incoherencies.
Is this based on mtime or ctime? If the former, would things could also
break if a user does, say, some stat(2), write(2), utimes(2) shenanigans?
So, my assumption is that if the mount option isn't there allowing this
then O_NOMTIME would be a no-op (as opposed to EPERM or something)... but
maybe that's not the right thing to do. Whatever we do there, though, I
suppose NFS would do the same thing?
sage
^ permalink raw reply
* Re: [PATCH 07/18] media controller: rename the tuner entity
From: Mauro Carvalho Chehab @ 2015-05-08 15:46 UTC (permalink / raw)
To: Hans Verkuil
Cc: Linux Media Mailing List, Mauro Carvalho Chehab, Jonathan Corbet,
Matthias Schwarzott, Antti Palosaari, Olli Salonen, Prabhakar Lad,
Sakari Ailus, Laurent Pinchart, linux-doc, linux-api
In-Reply-To: <554CC8E3.2030308@xs4all.nl>
Em Fri, 08 May 2015 16:32:03 +0200
Hans Verkuil <hverkuil@xs4all.nl> escreveu:
> On 05/08/2015 04:08 PM, Mauro Carvalho Chehab wrote:
> > Em Fri, 08 May 2015 15:21:39 +0200
> > Hans Verkuil <hverkuil@xs4all.nl> escreveu:
> >
> >> On 05/08/2015 02:57 PM, Mauro Carvalho Chehab wrote:
> >>> Em Fri, 08 May 2015 14:13:22 +0200
> >>> Hans Verkuil <hverkuil@xs4all.nl> escreveu:
> >>>
> >>>> On 05/08/2015 03:12 AM, Mauro Carvalho Chehab wrote:
> >>>>> Finally, let's rename the tuner entity. inside the media subsystem,
> >>>>> a tuner can be used by AM/FM radio, SDR radio, analog TV and digital TV.
> >>>>> It could even be used on other subsystems, like network, for wireless
> >>>>> devices.
> >>>>>
> >>>>> So, it is not constricted to V4L2 API, or to a subdev.
> >>>>>
> >>>>> Let's then rename it as:
> >>>>> MEDIA_ENT_T_V4L2_SUBDEV_TUNER -> MEDIA_ENT_T_TUNER
> >>>>
> >>>> See patch 04/18.
> >>>
> >>> Mapping the tuner as a V4L2_SUBDEV is plain wrong. We can't assume
> >>> that a tuner will always be mapped via V4L2 subdev API.
> >>
> >> True. Today we have subdevs that have no device node to control them, so
> >> in that case it would just be a SUBDEV entity. There are subdevs that make
> >> a v4l-subdev device node, so those can be V4L(2)_SUBDEV entities.
> >>
> >> The question is: what are your ideas for e.g. DVB-only tuners? Would they
> >> get a DVB-like device node? (so DTV_SUBDEV)
> >
> > I guess we may need DVB subdevs in the future. For now, I don't see
> > much usage.
> >
> >> Would hybrid tuners have two
> >> device nodes? One v4l-subdev, one dvb/dtv-subdev?
> >
> > No. A tuner is a tuner. The very same device can be used for analog or
> > digital TV. Ok, there are tuners that only work for digital TV (satellite
> > tuners, typically), because satellite requires a different tuning range,
> > and require an extra hardware to power up the satellite antena. So, on
> > most devices, the tuner is integrated with SEC.
> >
> > In any case, I don't see any reason why artificially one piece of hardware
> > component (tuner) into one subdevice entity per API.
> >
> > What it might make sense in the future is to have some DVB-specific ioctls
> > for a hybrid tuner, in order to allow adjusting its internal filters to
> > an specific digital TV standard.
> >
> >> Just curious what your thoughts are.
> >>
> >> Brainstorming:
> >>
> >> It might be better to map each device node to an entity and each hardware
> >> component (tuner, DMA engine) to an entity, and avoid this mixing of
> >> hw entity vs device node entity.
> >
> > Ok, but then we need to properly define the namespaces for HW and for
> > Linux API components.
> >
> > So, we would have a namespace like:
> >
> > - ENT_T_DEVNODE_DVB_(FE|CA|NET|DEMUX|DVR) for the DVB API device nodes;
> > - ENT_T_DEVNODE_V4L for the radio/swradio/video/vbi devnodes, or,
> > alternatively, use ENT_T_DEVNODE_V4L_(RADIO|SWRADIO|VIDEO|VBI);
> > - ENT_T_HW_(TUNER|CAM_SENSOR|ATV_DEMOD|DTV_DEMOD|...) for the
> > hardware components.
> >
> > In other words, the namespace would actually define two subtypes:
> > - devnodes;
> > - hardware components
> >
> > There's one advantage on this strategy: it is easier to keep backward
> > compatibility, IMHO, as we'll be preserving 1 << 16 for device nodes
>
> Right, that will work.
>
> > and 2 << 16 for hardware components.
>
> This remains problematic since I believe this should be done as a list
> of properties. Instead the entity type would be ENT_T_HW (if it represents
> actual hardware), ENT_T_SW (if it is a software implementation, for example
> for the DVB demux if there is no HW demux), and perhaps ENT_T_IP for
> representing IP blocks (not sure about this one).
I don't think we should distinguish IP block from HW. I would also use
a different namespace for SW implementation like the DVB demux or
for API devnodes.
Perhaps:
ENT_T_API - for non-subdev API device nodes;
ENT_T_HW - for hardware/firmware/IP blocks;
ENT_T_SW - for software emulation blocks like DVB demux
I would not map subdev API as a separate entity, as those should have
a 1:1 map to the entity. So, a property (or a flag) should be enough
to indicate that either a ENT_T_HW_ or a ENT_T_SW_ is exposed via a
subdev.
With regards to the API mapping, I would have a set of properties that
would be grouping V4L2, subdev, DVB, ... parts of the API.
So, for example, if the entity that represents a video devnode
controls the tuner, the analog demod and the DMA engine, such
entity would have 3 properties:
- ENT_PROP_V4L2_TUNER
- ENT_PROP_V4L2_ATV_DEMOD
- ENT_PROP_V4L2_DMA
> And the properties will tell what functions it supports.
Yes, but, as the API right now groups the subtypes, it would
be better to keep it splitted, not for userspace to use this as
subtype. Just as a matter of keeping things well organized.
I mean, if you take a look at the resulting namespace after all
those patches I wrote, the end result is:
#define MEDIA_ENT_T_UNKNOWN 0
#define MEDIA_ENT_T_AV_DMA (((1 << 16)) + 1)
#define MEDIA_ENT_T_DTV_DEMOD (MEDIA_ENT_T_AV_DMA + 3)
#define MEDIA_ENT_T_DTV_DEMUX (MEDIA_ENT_T_AV_DMA + 4)
#define MEDIA_ENT_T_DTV_DVR (MEDIA_ENT_T_AV_DMA + 5)
#define MEDIA_ENT_T_DTV_CA (MEDIA_ENT_T_AV_DMA + 6)
#define MEDIA_ENT_T_DTV_NET (MEDIA_ENT_T_AV_DMA + 7)
#define MEDIA_ENT_T_CAM_SENSOR ((2 << 16) + 1)
#define MEDIA_ENT_T_CAM_FLASH (MEDIA_ENT_T_CAM_SENSOR + 1)
#define MEDIA_ENT_T_CAM_LENS (MEDIA_ENT_T_CAM_SENSOR + 2)
#define MEDIA_ENT_T_ATV_DECODER (MEDIA_ENT_T_CAM_SENSOR + 3)
#define MEDIA_ENT_T_TUNER (MEDIA_ENT_T_CAM_SENSOR + 4)
We cannot renumber them, or otherwise things will break. So, we'll have
to deal with at least 2 number spaces: 1 << 16 and 2 << 16.
If we use the 1<<16 number space for ENT_T_API and the 2<<16 for
ENT_T_HW, the above will keep fitting, and it will be easier to maintain,
as we'll be grouping by name.
> The existing 2 << 16 defines would only be used if the property list matches
> the original meaning of the define to keep it backwards compatible.
>
> >
> > Yet, we'll need to add an entity for the V4L2 hardware DMA (with makes
> > sense to me), and this might break backward compatibility if not done
> > well.
>
> I see this as a property as well, but otherwise I agree with this.
Yes, this could be a property.
> >
> > It should be said that, in such case, hardware components will then
> > mean not only V4L2-specific hardware (V4L2_SUBDEV_foo), but also DVB,
> > ALSA, ... components.
> >
> > So, we'll still need a way to identify what of those components are
> > V4L2 subdevs, probably using the properties API.
>
> Why? A hardware component that can be controlled via a v4l-subdev node
> would be linked to an entity for that v4l-subdev node. That's an API
> entity. The whole 'is this a v4l-subdev' question disappears, since that
> is now no longer relevant. Instead you will have an ENT_T_DEVNODE_V4L_SUBDEV
> entity linked to the hw entity.
I see V4L_SUBDEV as a set of properties, specifying exactly what group
of subdev ioctls controls an specific hardware entity.
> Even a radio device would fit cleanly into this: the tuner entity simply
> has a link to a radio device node entity.
Yes.
> Hmm, does this also solve the control vs DMA issue? If a DEVNODE entity
> is hooked up to an entity with the DMA functionality, then you can stream,
> otherwise it is just for control.
Probably yes.
> I'm not sure if this is always true, though. Of course, we can also just
> add the streaming/dma property to the DEVNODE entity as well.
>
> > If you all agree with that, I'll respin the patch series to map the
> > entities like that.
>
> We can interpret the existing ENT_T_HW_TUNER etc. as shorthand for a
> property while the property API isn't there yet (we need that anyway
> for backwards compat).
I do prefer using a switch() like I did on patch 10/18, specifying the
entities that belong to subdev. After we add the property API, we can
convert is_media_entity_v4l2_subdev and is_media_entity_av_dma (with
should likely be renamed to is_media_entity_v4l_dma) to use the
properties API.
I don't think we'll need to create a separate number space for the new
entities in order to be backward-compatible.
>
> So we would need to add a ENT_T_HW_DMA as well (to be replaced by a
> property later).
Let's add this directly when we add the Property API.
> Basically I see the 1 << 16 range as device node types, the 2 << 16 range
> as shorthands for what should be properties (this really defined functions
> and entities can combine multiple functions), and we would need to have a
> new range (4 << 16) for non-DEVNODE entity types. Although we could keep
> it in range 1 << 16 as well, but I think it might make sense to keep it
> separate.
Keeping a backward-compatible separate range would create messy graphs.
I don't think we should do it.
>
> And there you would get ENT_T_HW/SW/IP (not sure about the last one). And
> perhaps FPGA.
>
> Again, just brainstorming here.
>
> Regards,
>
> Hans
^ permalink raw reply
* Re: [RFC PATCH 00/11] drm/i915: Expose OA metrics via perf PMU
From: Peter Zijlstra @ 2015-05-08 16:21 UTC (permalink / raw)
To: Robert Bragg
Cc: dri-devel, linux-api, intel-gfx, linux-kernel, Ingo Molnar,
Paul Mackerras, Arnaldo Carvalho de Melo, Daniel Vetter
In-Reply-To: <1431008154-6833-1-git-send-email-robert@sixbynine.org>
So I've not yet went through the entire series; but I'm wondering if its
at all possible to re-use some of this work:
lkml.kernel.org/r/1428453299-19121-1-git-send-email-sukadev@linux.vnet.ibm.com
That's for a Power8 HV call that can basically return an array of
values; which on a superficial level sounds a bit like what this GPU
hardware does.
Let me read more of this..
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply
* Re: [RFC PATCH 00/11] drm/i915: Expose OA metrics via perf PMU
From: Peter Zijlstra @ 2015-05-08 16:24 UTC (permalink / raw)
To: Robert Bragg
Cc: intel-gfx-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW, Daniel Vetter,
Jani Nikula, David Airlie, Paul Mackerras, Ingo Molnar,
Arnaldo Carvalho de Melo, linux-kernel-u79uwXL29TY76Z2rM5mHXA,
dri-devel-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW,
linux-api-u79uwXL29TY76Z2rM5mHXA
In-Reply-To: <1431008154-6833-1-git-send-email-robert-St23OQVBDYPNLxjTenLetw@public.gmane.org>
On Thu, May 07, 2015 at 03:15:43PM +0100, Robert Bragg wrote:
> I've changed the uapi for configuring the i915_oa specific attributes
> when calling perf_event_open(2) whereby instead of cramming lots of
> bitfields into the perf_event_attr config members, I'm now
> daisy-chaining a drm_i915_oa_event_attr_t structure off of a single
> config member that's extensible and validated in the same way as the
> perf_event_attr struct. I've found this much nicer to work with while
> being neatly extensible too.
This worries me a bit.. is there more background for this?
^ permalink raw reply
* Re: [PATCH v3 09/12] KVM: arm64: guest debug, HW assisted debug support
From: Christoffer Dall @ 2015-05-08 16:32 UTC (permalink / raw)
To: Alex Bennée
Cc: kvm-u79uwXL29TY76Z2rM5mHXA,
linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
kvmarm-FPEHb7Xf0XXUo1n7N8X6UoWGPAHP3yOg, marc.zyngier-5wv7dgnIgG8,
peter.maydell-QSEj5FYQhm4dnm+yROfE0A, agraf-l3A5Bk7waGM,
drjones-H+wXaHxf7aLQT0dZR+AlfA, pbonzini-H+wXaHxf7aLQT0dZR+AlfA,
zhichao.huang-QSEj5FYQhm4dnm+yROfE0A,
jan.kiszka-kv7WeFo6aLtBDgjK7y7TUQ,
dahi-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8,
r65777-KZfg59tc24xl57MIdRCFDg, bp-l3A5Bk7waGM, Gleb Natapov,
Jonathan Corbet, Russell King, Catalin Marinas, Will Deacon,
Ard Biesheuvel, Lorenzo Pieralisi, Andre Przywara,
Richard Weinberger, Peter Zijlstra, Ingo Molnar, AKASHI Takahiro,
open list:DOCUMENTATION, open list
In-Reply-To: <1430989647-22501-2-git-send-email-alex.bennee-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org>
On Thu, May 07, 2015 at 10:07:12AM +0100, Alex Bennée wrote:
> This adds support for userspace to control the HW debug registers for
> guest debug. In the debug ioctl we copy the IMPDEF defined number of
> registers into a new register set called host_debug_state. There is now
> a new vcpu parameter called debug_ptr which selects which register set
> is to copied into the real registers when world switch occurs.
>
> I've moved some helper functions into the hw_breakpoint.h header for
> re-use.
>
> As with single step we need to tweak the guest registers to enable the
> exceptions so we need to save and restore those bits.
>
> Two new capabilities have been added to the KVM_EXTENSION ioctl to allow
> userspace to query the number of hardware break and watch points
> available on the host hardware.
>
> Signed-off-by: Alex Bennée <alex.bennee-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org>
>
> ---
> v2
> - switched to C setup
> - replace host debug registers directly into context
> - minor tweak to api docs
> - setup right register for debug
> - add FAR_EL2 to debug exit structure
> - add support for trapping debug register access
> v3
> - remove stray trace statement
> - fix spacing around operators (various)
> - clean-up usage of trap_debug
> - introduce debug_ptr, replace excessive memcpy stuff
> - don't use memcpy in ioctl, just assign
> - update cap ioctl documentation
> - reword a number comments
> - rename host_debug_state->external_debug_state
>
> diff --git a/Documentation/virtual/kvm/api.txt b/Documentation/virtual/kvm/api.txt
> index 5ef937c..419f7a8 100644
> --- a/Documentation/virtual/kvm/api.txt
> +++ b/Documentation/virtual/kvm/api.txt
> @@ -2668,7 +2668,7 @@ The top 16 bits of the control field are architecture specific control
> flags which can include the following:
>
> - KVM_GUESTDBG_USE_SW_BP: using software breakpoints [x86, arm64]
> - - KVM_GUESTDBG_USE_HW_BP: using hardware breakpoints [x86, s390]
> + - KVM_GUESTDBG_USE_HW_BP: using hardware breakpoints [x86, s390, arm64]
> - KVM_GUESTDBG_INJECT_DB: inject DB type exception [x86]
> - KVM_GUESTDBG_INJECT_BP: inject BP type exception [x86]
> - KVM_GUESTDBG_EXIT_PENDING: trigger an immediate guest exit [s390]
> @@ -2683,6 +2683,11 @@ updated to the correct (supplied) values.
> The second part of the structure is architecture specific and
> typically contains a set of debug registers.
>
> +For arm64 the number of debug registers is implementation defined and
> +can be determined by querying the KVM_CAP_GUEST_DEBUG_HW_BPS and
> +KVM_CAP_GUEST_DEBUG_HW_WPS capabilities which returns a +ve number
s/returns/return/
s/+ve/positive/
> +indicating the number of supported registers.
> +
> When debug events exit the main run loop with the reason
> KVM_EXIT_DEBUG with the kvm_debug_exit_arch part of the kvm_run
> structure containing architecture specific debug information.
> diff --git a/arch/arm/kvm/arm.c b/arch/arm/kvm/arm.c
> index 9b3ed6d..2920185 100644
> --- a/arch/arm/kvm/arm.c
> +++ b/arch/arm/kvm/arm.c
> @@ -279,6 +279,10 @@ int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu)
> /* Set up the timer */
> kvm_timer_vcpu_init(vcpu);
>
> + /* Set the debug registers to be the guests */
> + vcpu->arch.debug_ptr = (struct kvm_guest_debug_arch *)
> + &vcpu_sys_reg(vcpu, DBGBCR0_EL1);
> +
yikes, I don't like this cast, how bad is it to get rid of the debug
registers in the sys_regs array ?
Also, pretty sure this is part of the breakage for the 32-bit build...
> return 0;
> }
>
> @@ -304,6 +308,7 @@ void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu)
>
> #define KVM_GUESTDBG_VALID_MASK (KVM_GUESTDBG_ENABLE | \
> KVM_GUESTDBG_USE_SW_BP | \
> + KVM_GUESTDBG_USE_HW_BP | \
> KVM_GUESTDBG_SINGLESTEP)
>
> /**
> @@ -324,6 +329,12 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
>
> if (dbg->control & KVM_GUESTDBG_ENABLE) {
> vcpu->guest_debug = dbg->control;
> +
> + /* Hardware assisted Break and Watch points */
> + if (vcpu->guest_debug & KVM_GUESTDBG_USE_HW_BP) {
is this only breakpoints or breakpoints and watch points?
> + vcpu->arch.external_debug_state = dbg->arch;
> + }
> +
> } else {
> /* If not enabled clear all flags */
> vcpu->guest_debug = 0;
> diff --git a/arch/arm64/include/asm/hw_breakpoint.h b/arch/arm64/include/asm/hw_breakpoint.h
> index 52b484b..c450552 100644
> --- a/arch/arm64/include/asm/hw_breakpoint.h
> +++ b/arch/arm64/include/asm/hw_breakpoint.h
> @@ -130,6 +130,18 @@ static inline void ptrace_hw_copy_thread(struct task_struct *task)
> }
> #endif
>
> +/* Determine number of BRP registers available. */
> +static inline int get_num_brps(void)
> +{
> + return ((read_cpuid(ID_AA64DFR0_EL1) >> 12) & 0xf) + 1;
> +}
> +
> +/* Determine number of WRP registers available. */
> +static inline int get_num_wrps(void)
> +{
> + return ((read_cpuid(ID_AA64DFR0_EL1) >> 20) & 0xf) + 1;
> +}
> +
I will need an ack from Catalin/Will to merge this. It may be better to
move these functions in a separate patch.
> extern struct pmu perf_ops_bp;
>
> #endif /* __KERNEL__ */
> diff --git a/arch/arm64/include/asm/kvm_host.h b/arch/arm64/include/asm/kvm_host.h
> index b60fa7a..a44fb32 100644
> --- a/arch/arm64/include/asm/kvm_host.h
> +++ b/arch/arm64/include/asm/kvm_host.h
> @@ -108,9 +108,18 @@ struct kvm_vcpu_arch {
> /* Exception Information */
> struct kvm_vcpu_fault_info fault;
>
> - /* Debug state */
> + /* Guest debug state */
> u64 debug_flags;
>
> + /*
> + * For debugging the guest we need to keep a set of debug
> + * registers which can override the guests own debug state
s/guests/guest's/
> + * while being used. These are set via the KVM_SET_GUEST_DEBUG
> + * ioctl.
> + */
> + struct kvm_guest_debug_arch *debug_ptr;
> + struct kvm_guest_debug_arch external_debug_state;
> +
> /* Pointer to host CPU context */
> kvm_cpu_context_t *host_cpu_context;
>
> diff --git a/arch/arm64/include/uapi/asm/kvm.h b/arch/arm64/include/uapi/asm/kvm.h
> index 04957d7..98e82ef 100644
> --- a/arch/arm64/include/uapi/asm/kvm.h
> +++ b/arch/arm64/include/uapi/asm/kvm.h
> @@ -121,7 +121,7 @@ struct kvm_guest_debug_arch {
>
> struct kvm_debug_exit_arch {
> __u32 hsr;
> - __u64 far;
> + __u64 far; /* used for watchpoints */
seems strange to amend this now?
> };
>
> struct kvm_sync_regs {
> diff --git a/arch/arm64/kernel/asm-offsets.c b/arch/arm64/kernel/asm-offsets.c
> index ce7b7dd..671ab13 100644
> --- a/arch/arm64/kernel/asm-offsets.c
> +++ b/arch/arm64/kernel/asm-offsets.c
> @@ -116,6 +116,7 @@ int main(void)
> DEFINE(VCPU_FAR_EL2, offsetof(struct kvm_vcpu, arch.fault.far_el2));
> DEFINE(VCPU_HPFAR_EL2, offsetof(struct kvm_vcpu, arch.fault.hpfar_el2));
> DEFINE(VCPU_DEBUG_FLAGS, offsetof(struct kvm_vcpu, arch.debug_flags));
> + DEFINE(VCPU_DEBUG_PTR, offsetof(struct kvm_vcpu, arch.debug_ptr));
> DEFINE(DEBUG_BCR, offsetof(struct kvm_guest_debug_arch, dbg_bcr));
> DEFINE(DEBUG_BVR, offsetof(struct kvm_guest_debug_arch, dbg_bvr));
> DEFINE(DEBUG_WCR, offsetof(struct kvm_guest_debug_arch, dbg_wcr));
> diff --git a/arch/arm64/kernel/hw_breakpoint.c b/arch/arm64/kernel/hw_breakpoint.c
> index e7d934d..3a41bbf 100644
> --- a/arch/arm64/kernel/hw_breakpoint.c
> +++ b/arch/arm64/kernel/hw_breakpoint.c
> @@ -49,18 +49,6 @@ static DEFINE_PER_CPU(int, stepping_kernel_bp);
> static int core_num_brps;
> static int core_num_wrps;
>
> -/* Determine number of BRP registers available. */
> -static int get_num_brps(void)
> -{
> - return ((read_cpuid(ID_AA64DFR0_EL1) >> 12) & 0xf) + 1;
> -}
> -
> -/* Determine number of WRP registers available. */
> -static int get_num_wrps(void)
> -{
> - return ((read_cpuid(ID_AA64DFR0_EL1) >> 20) & 0xf) + 1;
> -}
> -
> int hw_breakpoint_slots(int type)
> {
> /*
> diff --git a/arch/arm64/kvm/debug.c b/arch/arm64/kvm/debug.c
> index 19346e8..1ab63dd 100644
> --- a/arch/arm64/kvm/debug.c
> +++ b/arch/arm64/kvm/debug.c
> @@ -99,12 +99,6 @@ void kvm_arm_setup_debug(struct kvm_vcpu *vcpu)
> MDCR_EL2_TDRA |
> MDCR_EL2_TDOSA);
>
> - /* Trap on access to debug registers? */
> - if (trap_debug)
> - vcpu->arch.mdcr_el2 |= MDCR_EL2_TDA;
> - else
> - vcpu->arch.mdcr_el2 &= ~MDCR_EL2_TDA;
> -
> /* Is Guest debugging in effect? */
> if (vcpu->guest_debug) {
> vcpu->arch.mdcr_el2 |= MDCR_EL2_TDE;
> @@ -128,14 +122,54 @@ void kvm_arm_setup_debug(struct kvm_vcpu *vcpu)
> vcpu_sys_reg(vcpu, MDSCR_EL1) &= ~DBG_MDSCR_SS;
> }
>
> + /*
> + * HW Break/Watch points
> + *
> + * We simply switch the debug_ptr to point to our new
> + * external_debug_state which has been populated by the
> + * debug ioctl. The existing KVM_ARM64_DEBUG_DIRTY
> + * mechanism ensures the registers are updated on the
> + * world switch.
> + */
> + if (vcpu->guest_debug & KVM_GUESTDBG_USE_HW_BP) {
> +
> + vcpu_sys_reg(vcpu, MDSCR_EL1) |=
> + (DBG_MDSCR_KDE | DBG_MDSCR_MDE);
Why do we need to set these two bits? Is it obvious or does it deserve
a comment?
> +
> + vcpu->arch.debug_ptr = &vcpu->arch.external_debug_state;
> + vcpu->arch.debug_flags |= KVM_ARM64_DEBUG_DIRTY;
> + trap_debug = true;
> + }
> +
> } else {
> /* Debug operations can go straight to the guest */
> vcpu->arch.mdcr_el2 &= ~MDCR_EL2_TDE;
> }
> +
> + /*
> + * If the guest debug register state is dirty (the guest is
> + * actively accessing them), then we context-switch the
> + * registers in EL2. Otherwise, we trap-and-emulate all guest
> + * accesses to them.
> + */
I think this comment now feels strange, because it was explaining why we
would set the trap_debug variable when the dirty flag was set, but the
code just sets TDA when trap_debug is set. So you should either move
this comment to the top of the function and have it above a separate
line that sets trap_debug based on KVM_ARM64_DEBUG_DIRTY (instead of
initializing at declaration), or you should explain which conditions set
trap_debug (guest is using the regs or we are debugging the guest), or
just get rid of the comment.
> + if (trap_debug)
> + vcpu->arch.mdcr_el2 |= MDCR_EL2_TDA;
> + else
> + vcpu->arch.mdcr_el2 &= ~MDCR_EL2_TDA;
still don't need the else.
> }
>
> void kvm_arm_clear_debug(struct kvm_vcpu *vcpu)
> {
> - if (vcpu->guest_debug)
> + if (vcpu->guest_debug) {
> restore_guest_debug_regs(vcpu);
> +
> + /*
> + * If we were using HW debug we need to restore the
> + * debug_ptr to the guest debug state.
> + */
> + if (vcpu->guest_debug & KVM_GUESTDBG_USE_HW_BP) {
> + vcpu->arch.debug_ptr = (struct kvm_guest_debug_arch *)
> + &vcpu_sys_reg(vcpu, DBGBCR0_EL1);
> + }
I would find it easier to follow the code if you only configure the
debug_ptr in kvm_arm_setup_debug() because it feels like you're setting
up state here which will not be used before in a very long time (after
handle_exit, exit to userspace etc.).
> + }
> }
> diff --git a/arch/arm64/kvm/handle_exit.c b/arch/arm64/kvm/handle_exit.c
> index e9de13e..68a0759 100644
> --- a/arch/arm64/kvm/handle_exit.c
> +++ b/arch/arm64/kvm/handle_exit.c
> @@ -103,7 +103,11 @@ static int kvm_handle_guest_debug(struct kvm_vcpu *vcpu, struct kvm_run *run)
> run->debug.arch.hsr = hsr;
>
> switch (hsr >> ESR_ELx_EC_SHIFT) {
> + case ESR_ELx_EC_WATCHPT_LOW:
> + run->debug.arch.far = vcpu->arch.fault.far_el2;
> + /* fall through */
> case ESR_ELx_EC_SOFTSTP_LOW:
> + case ESR_ELx_EC_BREAKPT_LOW:
> case ESR_ELx_EC_BKPT32:
> case ESR_ELx_EC_BRK64:
> break;
> @@ -132,6 +136,8 @@ static exit_handle_fn arm_exit_handlers[] = {
> [ESR_ELx_EC_IABT_LOW] = kvm_handle_guest_abort,
> [ESR_ELx_EC_DABT_LOW] = kvm_handle_guest_abort,
> [ESR_ELx_EC_SOFTSTP_LOW]= kvm_handle_guest_debug,
> + [ESR_ELx_EC_WATCHPT_LOW]= kvm_handle_guest_debug,
> + [ESR_ELx_EC_BREAKPT_LOW]= kvm_handle_guest_debug,
> [ESR_ELx_EC_BKPT32] = kvm_handle_guest_debug,
> [ESR_ELx_EC_BRK64] = kvm_handle_guest_debug,
> };
> diff --git a/arch/arm64/kvm/hyp.S b/arch/arm64/kvm/hyp.S
> index dd51fb1..921d248 100644
> --- a/arch/arm64/kvm/hyp.S
> +++ b/arch/arm64/kvm/hyp.S
> @@ -706,7 +706,8 @@ ENTRY(__kvm_vcpu_run)
> bl __restore_fpsimd
>
> skip_debug_state x3, 1f
> - add x3, x2, #CPU_SYSREG_OFFSET(DBGBCR0_EL1)
> + ldr x3, [x0, #VCPU_DEBUG_PTR]
> + kern_hyp_va x3
> bl __restore_debug
> 1:
> restore_guest_32bit_state
> @@ -727,7 +728,8 @@ __kvm_vcpu_return:
> bl __save_sysregs
>
> skip_debug_state x3, 1f
> - add x3, x2, #CPU_SYSREG_OFFSET(DBGBCR0_EL1)
> + ldr x3, [x0, #VCPU_DEBUG_PTR]
> + kern_hyp_va x3
> bl __save_debug
> 1:
> save_guest_32bit_state
> diff --git a/arch/arm64/kvm/reset.c b/arch/arm64/kvm/reset.c
> index 0b43265..21d5a62 100644
> --- a/arch/arm64/kvm/reset.c
> +++ b/arch/arm64/kvm/reset.c
> @@ -56,6 +56,12 @@ static bool cpu_has_32bit_el1(void)
> return !!(pfr0 & 0x20);
> }
>
> +/**
> + * kvm_arch_dev_ioctl_check_extension
> + *
> + * We currently assume that the number of HW registers is uniform
> + * across all CPUs (see cpuinfo_sanity_check).
> + */
> int kvm_arch_dev_ioctl_check_extension(long ext)
> {
> int r;
> @@ -64,6 +70,12 @@ int kvm_arch_dev_ioctl_check_extension(long ext)
> case KVM_CAP_ARM_EL1_32BIT:
> r = cpu_has_32bit_el1();
> break;
> + case KVM_CAP_GUEST_DEBUG_HW_BPS:
> + r = get_num_brps();
> + break;
> + case KVM_CAP_GUEST_DEBUG_HW_WPS:
> + r = get_num_wrps();
> + break;
> default:
> r = 0;
> }
> diff --git a/include/uapi/linux/kvm.h b/include/uapi/linux/kvm.h
> index 3b6252e..923c2aa 100644
> --- a/include/uapi/linux/kvm.h
> +++ b/include/uapi/linux/kvm.h
> @@ -825,6 +825,8 @@ struct kvm_ppc_smmu_info {
> #define KVM_CAP_S390_INJECT_IRQ 113
> #define KVM_CAP_S390_IRQ_STATE 114
> #define KVM_CAP_PPC_HWRNG 115
> +#define KVM_CAP_GUEST_DEBUG_HW_BPS 116
> +#define KVM_CAP_GUEST_DEBUG_HW_WPS 117
>
> #ifdef KVM_CAP_IRQ_ROUTING
>
> --
> 2.3.5
>
Thanks,
-Christoffer
^ permalink raw reply
* Re: [PATCH RFC] vfs: add a O_NOMTIME flag
From: Zach Brown @ 2015-05-08 17:11 UTC (permalink / raw)
To: Sage Weil
Cc: Dave Chinner, Alexander Viro,
linux-fsdevel-u79uwXL29TY76Z2rM5mHXA,
linux-kernel-u79uwXL29TY76Z2rM5mHXA,
linux-api-u79uwXL29TY76Z2rM5mHXA
In-Reply-To: <alpine.DEB.2.00.1505071752520.28239-vIokxiIdD2AQNTJnQDzGJqxOck334EZe@public.gmane.org>
On Thu, May 07, 2015 at 06:01:23PM -0700, Sage Weil wrote:
> On Thu, 7 May 2015, Zach Brown wrote:
> > On Thu, May 07, 2015 at 10:26:17AM +1000, Dave Chinner wrote:
> > > On Wed, May 06, 2015 at 03:00:12PM -0700, Zach Brown wrote:
> > > > The criteria for using O_NOMTIME is the same as for using O_NOATIME:
> > > > owning the file or having the CAP_FOWNER capability. If we're not
> > > > comfortable allowing owners to prevent mtime/ctime updates then we
> > > > should add a tunable to allow O_NOMTIME. Maybe a mount option?
> > >
> > > I dislike "turn off safety for performance" options because Joe
> > > SpeedRacer will always select performance over safety.
> >
> > Well, for ceph there's no safety concern. They never use cmtime in
> > these files.
> >
> > So are you suggesting not implementing this and making them rework their
> > IO paths to avoid the fs maintaining mtime so that we don't give Joe
> > Speedracer more rope? Or are we talking about adding some speed bumps
> > that ceph can flip on that might give Joe Speedracer pause?
>
> I think this is the fundamental question: who do we give the ammunition
> to, the user or app writer, or the sysadmin?
Yeah, I think this is right. Dave doesn't want the possibility of it
bleeding in to installations through irresponsible default use in apps
without explicit buy-in from the people responsible for the backups.
> [...]
>
> Or, we can be conservative and require a mount option so that the admin
> has to explicitly allow behavior that might break some existing
> assumptions about mtime/ctime ('-o user_noatime' I guess?).
>
> I'm happy either way, so long as in the end an unprivileged ceph daemon
> avoids the useless work. In our case we always own the entire mount/disk,
> so a mount option is just fine.
It seems that the thread has headed towards responding to my suggestion
of a possible mount option with an enthusiastic "yes, please, no
surprises."
So I'll try that.
- z
^ permalink raw reply
* [PATCH 0/6] support "dataplane" mode for nohz_full
From: Chris Metcalf @ 2015-05-08 17:58 UTC (permalink / raw)
To: Gilad Ben Yossef, Steven Rostedt, Ingo Molnar, Peter Zijlstra,
Andrew Morton, Rik van Riel, Tejun Heo, Frederic Weisbecker,
Thomas Gleixner, Paul E. McKenney, Christoph Lameter,
Srivatsa S. Bhat, linux-doc, linux-api, linux-kernel
Cc: Chris Metcalf
The existing nohz_full mode does a nice job of suppressing extraneous
kernel interrupts for cores that desire it. However, there is a need
for a more deterministic mode that rigorously disallows kernel
interrupts, even at a higher cost in user/kernel transition time:
for example, high-speed networking applications running userspace
drivers that will drop packets if they are ever interrupted.
These changes attempt to provide an initial draft of such a framework;
the changes do not add any overhead to the usual non-nohz_full mode,
and only very small overhead to the typical nohz_full mode. A prctl()
option (PR_SET_DATAPLANE) is added to control whether processes have
requested this stricter semantics, and within that prctl() option we
provide a number of different bits for more precise control.
Additionally, we add a new command-line boot argument to facilitate
debugging where unexpected interrupts are being delivered from.
Code that is conceptually similar has been in use in Tilera's
Multicore Development Environment since 2008, known as Zero-Overhead
Linux, and has seen wide adoption by a range of customers. This patch
series represents the first serious attempt to upstream that
functionality. Although the current state of the kernel isn't quite
ready to run with absolutely no kernel interrupts (for example,
workqueues on dataplane cores still remain to be dealt with), this
patch series provides a way to make dynamic tradeoffs between avoiding
kernel interrupts on the one hand, and making voluntary calls in and
out of the kernel more expensive, for tasks that want it.
The series (based currently on my arch/tile master tree for 4.2,
in turn based on 4.1-rc1) is available at:
git://git.kernel.org/pub/scm/linux/kernel/git/cmetcalf/linux-tile.git dataplane
Chris Metcalf (6):
nohz_full: add support for "dataplane" mode
nohz: dataplane: allow tick to be fully disabled for dataplane
dataplane nohz: run softirqs synchronously on user entry
nohz: support PR_DATAPLANE_QUIESCE
nohz: support PR_DATAPLANE_STRICT mode
nohz: add dataplane_debug boot flag
Documentation/kernel-parameters.txt | 6 ++
arch/tile/mm/homecache.c | 5 +-
include/linux/sched.h | 3 +
include/linux/tick.h | 12 ++++
include/uapi/linux/prctl.h | 8 +++
kernel/context_tracking.c | 3 +
kernel/irq_work.c | 4 +-
kernel/sched/core.c | 18 ++++++
kernel/signal.c | 5 ++
kernel/smp.c | 4 ++
kernel/softirq.c | 15 ++++-
kernel/sys.c | 8 +++
kernel/time/tick-sched.c | 112 +++++++++++++++++++++++++++++++++++-
13 files changed, 198 insertions(+), 5 deletions(-)
--
2.1.2
^ permalink raw reply
* [PATCH 1/6] nohz_full: add support for "dataplane" mode
From: Chris Metcalf @ 2015-05-08 17:58 UTC (permalink / raw)
To: Gilad Ben Yossef, Steven Rostedt, Ingo Molnar, Peter Zijlstra,
Andrew Morton, Rik van Riel, Tejun Heo, Frederic Weisbecker,
Thomas Gleixner, Paul E. McKenney, Christoph Lameter,
Srivatsa S. Bhat, linux-doc, linux-api, linux-kernel
Cc: Chris Metcalf
In-Reply-To: <1431107927-13998-1-git-send-email-cmetcalf@ezchip.com>
The existing nohz_full mode makes tradeoffs to minimize userspace
interruptions while still attempting to avoid overheads in the
kernel entry/exit path, to provide 100% kernel semantics, etc.
However, some applications require a stronger commitment from the
kernel to avoid interruptions, in particular userspace device
driver style applications, such as high-speed networking code.
This change introduces a framework to allow applications to elect
to have the stronger semantics as needed, specifying
prctl(PR_SET_DATAPLANE, PR_DATAPLANE_ENABLE) to do so.
Subsequent commits will add additional flags and additional
semantics.
The dataplane state is indicated by setting a new task struct
field, dataplane_flags, to the value passed by prctl(). When the
_ENABLE bit is set for a task, and it is returning to userspace
on a nohz_full core, it calls the new tick_nohz_dataplane_enter()
routine to take additional actions to help the task avoid being
interrupted in the future.
For this first patch, the only action taken is to call
lru_add_drain() to prevent being interrupted by a subsequent
lru_add_drain_all() call on another core.
Signed-off-by: Chris Metcalf <cmetcalf@ezchip.com>
---
include/linux/sched.h | 3 +++
include/linux/tick.h | 10 ++++++++++
include/uapi/linux/prctl.h | 5 +++++
kernel/context_tracking.c | 3 +++
kernel/sys.c | 8 ++++++++
kernel/time/tick-sched.c | 13 +++++++++++++
6 files changed, 42 insertions(+)
diff --git a/include/linux/sched.h b/include/linux/sched.h
index 8222ae40ecb0..3680aa07c9ea 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -1732,6 +1732,9 @@ struct task_struct {
#ifdef CONFIG_DEBUG_ATOMIC_SLEEP
unsigned long task_state_change;
#endif
+#ifdef CONFIG_NO_HZ_FULL
+ unsigned int dataplane_flags;
+#endif
};
/* Future-safe accessor for struct task_struct's cpus_allowed. */
diff --git a/include/linux/tick.h b/include/linux/tick.h
index f8492da57ad3..d191cda9b71a 100644
--- a/include/linux/tick.h
+++ b/include/linux/tick.h
@@ -10,6 +10,7 @@
#include <linux/context_tracking_state.h>
#include <linux/cpumask.h>
#include <linux/sched.h>
+#include <linux/prctl.h>
#ifdef CONFIG_GENERIC_CLOCKEVENTS
extern void __init tick_init(void);
@@ -134,11 +135,18 @@ static inline bool tick_nohz_full_cpu(int cpu)
return cpumask_test_cpu(cpu, tick_nohz_full_mask);
}
+static inline bool tick_nohz_is_dataplane(void)
+{
+ return tick_nohz_full_cpu(smp_processor_id()) &&
+ (current->dataplane_flags & PR_DATAPLANE_ENABLE);
+}
+
extern void __tick_nohz_full_check(void);
extern void tick_nohz_full_kick(void);
extern void tick_nohz_full_kick_cpu(int cpu);
extern void tick_nohz_full_kick_all(void);
extern void __tick_nohz_task_switch(struct task_struct *tsk);
+extern void tick_nohz_dataplane_enter(void);
#else
static inline bool tick_nohz_full_enabled(void) { return false; }
static inline bool tick_nohz_full_cpu(int cpu) { return false; }
@@ -147,6 +155,8 @@ static inline void tick_nohz_full_kick_cpu(int cpu) { }
static inline void tick_nohz_full_kick(void) { }
static inline void tick_nohz_full_kick_all(void) { }
static inline void __tick_nohz_task_switch(struct task_struct *tsk) { }
+static inline bool tick_nohz_is_dataplane(void) { return false; }
+static inline void tick_nohz_dataplane_enter(void) { }
#endif
static inline bool is_housekeeping_cpu(int cpu)
diff --git a/include/uapi/linux/prctl.h b/include/uapi/linux/prctl.h
index 31891d9535e2..1aa8fa8a8b05 100644
--- a/include/uapi/linux/prctl.h
+++ b/include/uapi/linux/prctl.h
@@ -190,4 +190,9 @@ struct prctl_mm_map {
# define PR_FP_MODE_FR (1 << 0) /* 64b FP registers */
# define PR_FP_MODE_FRE (1 << 1) /* 32b compatibility */
+/* Enable/disable or query dataplane mode for NO_HZ_FULL kernels. */
+#define PR_SET_DATAPLANE 47
+#define PR_GET_DATAPLANE 48
+# define PR_DATAPLANE_ENABLE (1 << 0)
+
#endif /* _LINUX_PRCTL_H */
diff --git a/kernel/context_tracking.c b/kernel/context_tracking.c
index 72d59a1a6eb6..dd6bdd6197b6 100644
--- a/kernel/context_tracking.c
+++ b/kernel/context_tracking.c
@@ -20,6 +20,7 @@
#include <linux/hardirq.h>
#include <linux/export.h>
#include <linux/kprobes.h>
+#include <linux/tick.h>
#define CREATE_TRACE_POINTS
#include <trace/events/context_tracking.h>
@@ -85,6 +86,8 @@ void context_tracking_enter(enum ctx_state state)
* on the tick.
*/
if (state == CONTEXT_USER) {
+ if (tick_nohz_is_dataplane())
+ tick_nohz_dataplane_enter();
trace_user_enter(0);
vtime_user_enter(current);
}
diff --git a/kernel/sys.c b/kernel/sys.c
index a4e372b798a5..930b750aefde 100644
--- a/kernel/sys.c
+++ b/kernel/sys.c
@@ -2243,6 +2243,14 @@ SYSCALL_DEFINE5(prctl, int, option, unsigned long, arg2, unsigned long, arg3,
case PR_GET_FP_MODE:
error = GET_FP_MODE(me);
break;
+#ifdef CONFIG_NO_HZ_FULL
+ case PR_SET_DATAPLANE:
+ me->dataplane_flags = arg2;
+ break;
+ case PR_GET_DATAPLANE:
+ error = me->dataplane_flags;
+ break;
+#endif
default:
error = -EINVAL;
break;
diff --git a/kernel/time/tick-sched.c b/kernel/time/tick-sched.c
index 914259128145..31c674719647 100644
--- a/kernel/time/tick-sched.c
+++ b/kernel/time/tick-sched.c
@@ -24,6 +24,7 @@
#include <linux/posix-timers.h>
#include <linux/perf_event.h>
#include <linux/context_tracking.h>
+#include <linux/swap.h>
#include <asm/irq_regs.h>
@@ -389,6 +390,18 @@ void __init tick_nohz_init(void)
pr_info("NO_HZ: Full dynticks CPUs: %*pbl.\n",
cpumask_pr_args(tick_nohz_full_mask));
}
+
+/*
+ * When returning to userspace on a nohz_full core after doing
+ * prctl(PR_DATAPLANE_SET,1), we come here and try more aggressively
+ * to prevent this core from being interrupted later.
+ */
+void tick_nohz_dataplane_enter(void)
+{
+ /* Drain the pagevecs to avoid unnecessary IPI flushes later. */
+ lru_add_drain();
+}
+
#endif
/*
--
2.1.2
^ permalink raw reply related
* [PATCH 4/6] nohz: support PR_DATAPLANE_QUIESCE
From: Chris Metcalf @ 2015-05-08 17:58 UTC (permalink / raw)
To: Gilad Ben Yossef, Steven Rostedt, Ingo Molnar, Peter Zijlstra,
Andrew Morton, Rik van Riel, Tejun Heo, Thomas Gleixner,
Frederic Weisbecker, Paul E. McKenney, Christoph Lameter,
Srivatsa S. Bhat, linux-doc, linux-api, linux-kernel
Cc: Chris Metcalf
In-Reply-To: <1431107927-13998-1-git-send-email-cmetcalf@ezchip.com>
This prctl() flag for PR_SET_DATAPLANE sets a mode that requires the
kernel to quiesce any pending timer interrupts prior to returning
to userspace. When running with this mode set, sys calls (and page
faults, etc.) can be inordinately slow. However, user applications
that want to guarantee that no unexpected interrupts will occur
(even if they call into the kernel) can set this flag to guarantee
that semantics.
Signed-off-by: Chris Metcalf <cmetcalf@ezchip.com>
---
include/uapi/linux/prctl.h | 1 +
kernel/time/tick-sched.c | 54 ++++++++++++++++++++++++++++++++++++++++++++++
2 files changed, 55 insertions(+)
diff --git a/include/uapi/linux/prctl.h b/include/uapi/linux/prctl.h
index 1aa8fa8a8b05..8b735651304a 100644
--- a/include/uapi/linux/prctl.h
+++ b/include/uapi/linux/prctl.h
@@ -194,5 +194,6 @@ struct prctl_mm_map {
#define PR_SET_DATAPLANE 47
#define PR_GET_DATAPLANE 48
# define PR_DATAPLANE_ENABLE (1 << 0)
+# define PR_DATAPLANE_QUIESCE (1 << 1)
#endif /* _LINUX_PRCTL_H */
diff --git a/kernel/time/tick-sched.c b/kernel/time/tick-sched.c
index fd0e6e5c931c..69d908c6cef8 100644
--- a/kernel/time/tick-sched.c
+++ b/kernel/time/tick-sched.c
@@ -392,6 +392,53 @@ void __init tick_nohz_init(void)
}
/*
+ * We normally return immediately to userspace.
+ *
+ * The PR_DATAPLANE_QUIESCE flag causes us to wait until no more
+ * interrupts are pending. Otherwise we nap with interrupts enabled
+ * and wait for the next interrupt to fire, then loop back and retry.
+ *
+ * Note that if you schedule two processes on the same core and both
+ * specify PR_DATAPLANE_QUIESCE, neither will ever leave the kernel,
+ * and one will have to be killed manually. Otherwise in situations
+ * where another process is in the runqueue on this cpu, this task
+ * will just wait for that other task to go idle before returning to
+ * user space.
+ */
+static void dataplane_quiesce(void)
+{
+ struct clock_event_device *dev =
+ __this_cpu_read(tick_cpu_device.evtdev);
+ struct task_struct *task = current;
+ unsigned long start = jiffies;
+ bool warned = false;
+
+ while (ACCESS_ONCE(dev->next_event.tv64) != KTIME_MAX) {
+ if (!warned && (jiffies - start) >= (5 * HZ)) {
+ pr_warn("%s/%d: cpu %d: dataplane task blocked for %ld jiffies\n",
+ task->comm, task->pid, smp_processor_id(),
+ (jiffies - start));
+ warned = true;
+ }
+ if (should_resched())
+ schedule();
+ if (test_thread_flag(TIF_SIGPENDING))
+ break;
+
+ /* Idle with interrupts enabled and wait for the tick. */
+ set_current_state(TASK_INTERRUPTIBLE);
+ arch_cpu_idle();
+ set_current_state(TASK_RUNNING);
+ }
+ if (warned) {
+ pr_warn("%s/%d: cpu %d: dataplane task unblocked after %ld jiffies\n",
+ task->comm, task->pid, smp_processor_id(),
+ (jiffies - start));
+ dump_stack();
+ }
+}
+
+/*
* When returning to userspace on a nohz_full core after doing
* prctl(PR_DATAPLANE_SET,1), we come here and try more aggressively
* to prevent this core from being interrupted later.
@@ -411,6 +458,13 @@ void tick_nohz_dataplane_enter(void)
lru_add_drain();
/*
+ * Quiesce any timer ticks if requested. On return from this
+ * function, no timer ticks are pending.
+ */
+ if ((current->dataplane_flags & PR_DATAPLANE_QUIESCE) != 0)
+ dataplane_quiesce();
+
+ /*
* Disable interrupts again since other code running in this
* function may have enabled them, and the caller expects
* interrupts to be disabled on return. Enabling them during
--
2.1.2
^ permalink raw reply related
* [PATCH 5/6] nohz: support PR_DATAPLANE_STRICT mode
From: Chris Metcalf @ 2015-05-08 17:58 UTC (permalink / raw)
To: Gilad Ben Yossef, Steven Rostedt, Ingo Molnar, Peter Zijlstra,
Andrew Morton, Rik van Riel, Tejun Heo, Thomas Gleixner,
Frederic Weisbecker, Paul E. McKenney, Christoph Lameter,
Srivatsa S. Bhat, linux-doc, linux-api, linux-kernel
Cc: Chris Metcalf
In-Reply-To: <1431107927-13998-1-git-send-email-cmetcalf@ezchip.com>
With QUIESCE mode, the task is in principle guaranteed not to be
interrupted by the kernel, but only if it behaves. In particular,
if it enters the kernel via system call, page fault, or any of
a number of other synchronous traps, it may be unexpectedly
exposed to long latencies. Add a simple flag that puts the process
into a state where any such kernel entry is fatal.
To allow the state to be entered and exited, we add an internal
bit to current->dataplane_flags that is set when prctl() sets the
flags. That way, when we are exiting the kernel after calling
prctl() to forbid future kernel exits, we don't get immediately
killed.
Signed-off-by: Chris Metcalf <cmetcalf@ezchip.com>
---
include/uapi/linux/prctl.h | 2 ++
kernel/sys.c | 2 +-
kernel/time/tick-sched.c | 17 +++++++++++++++++
3 files changed, 20 insertions(+), 1 deletion(-)
diff --git a/include/uapi/linux/prctl.h b/include/uapi/linux/prctl.h
index 8b735651304a..9cf79aa1e73f 100644
--- a/include/uapi/linux/prctl.h
+++ b/include/uapi/linux/prctl.h
@@ -195,5 +195,7 @@ struct prctl_mm_map {
#define PR_GET_DATAPLANE 48
# define PR_DATAPLANE_ENABLE (1 << 0)
# define PR_DATAPLANE_QUIESCE (1 << 1)
+# define PR_DATAPLANE_STRICT (1 << 2)
+# define PR_DATAPLANE_PRCTL (1U << 31) /* kernel internal */
#endif /* _LINUX_PRCTL_H */
diff --git a/kernel/sys.c b/kernel/sys.c
index 930b750aefde..8102433c9edd 100644
--- a/kernel/sys.c
+++ b/kernel/sys.c
@@ -2245,7 +2245,7 @@ SYSCALL_DEFINE5(prctl, int, option, unsigned long, arg2, unsigned long, arg3,
break;
#ifdef CONFIG_NO_HZ_FULL
case PR_SET_DATAPLANE:
- me->dataplane_flags = arg2;
+ me->dataplane_flags = arg2 | PR_DATAPLANE_PRCTL;
break;
case PR_GET_DATAPLANE:
error = me->dataplane_flags;
diff --git a/kernel/time/tick-sched.c b/kernel/time/tick-sched.c
index 69d908c6cef8..22ed0decb363 100644
--- a/kernel/time/tick-sched.c
+++ b/kernel/time/tick-sched.c
@@ -436,6 +436,20 @@ static void dataplane_quiesce(void)
(jiffies - start));
dump_stack();
}
+
+ /*
+ * Kill the process if it violates STRICT mode. Note that this
+ * code also results in killing the task if a kernel bug causes an
+ * irq to be delivered to this core.
+ */
+ if ((task->dataplane_flags & (PR_DATAPLANE_STRICT|PR_DATAPLANE_PRCTL))
+ == PR_DATAPLANE_STRICT) {
+ pr_warn("Dataplane STRICT mode violated; process killed.\n");
+ dump_stack();
+ task->dataplane_flags &= ~PR_DATAPLANE_QUIESCE;
+ local_irq_enable();
+ do_group_exit(SIGKILL);
+ }
}
/*
@@ -464,6 +478,9 @@ void tick_nohz_dataplane_enter(void)
if ((current->dataplane_flags & PR_DATAPLANE_QUIESCE) != 0)
dataplane_quiesce();
+ /* Clear the bit set by prctl() when it updates the flags. */
+ current->dataplane_flags &= ~PR_DATAPLANE_PRCTL;
+
/*
* Disable interrupts again since other code running in this
* function may have enabled them, and the caller expects
--
2.1.2
^ permalink raw reply related
* [PATCH 0/3] Allow user to request memory to be locked on page fault
From: Eric B Munson @ 2015-05-08 19:33 UTC (permalink / raw)
To: Andrew Morton
Cc: Eric B Munson, Shuah Khan, linux-alpha, linux-kernel, linux-mips,
linux-parisc, linuxppc-dev, sparclinux, linux-xtensa, linux-mm,
linux-arch, linux-api
mlock() allows a user to control page out of program memory, but this
comes at the cost of faulting in the entire mapping when it is
allocated. For large mappings where the entire area is not necessary
this is not ideal.
This series introduces new flags for mmap() and mlockall() that allow a
user to specify that the covered are should not be paged out, but only
after the memory has been used the first time.
The performance cost of these patches are minimal on the two benchmarks
I have tested (stream and kernbench).
Avg throughput in MB/s from stream using 1000000 element arrays
Test 4.1-rc2 4.1-rc2+lock-on-fault
Copy: 10,979.08 10,917.34
Scale: 11,094.45 11,023.01
Add: 12,487.29 12,388.65
Triad: 12,505.77 12,418.78
Kernbench optimal load
4.1-rc2 4.1-rc2+lock-on-fault
Elapsed Time 71.046 71.324
User Time 62.117 62.352
System Time 8.926 8.969
Context Switches 14531.9 14542.5
Sleeps 14935.9 14939
Eric B Munson (3):
Add flag to request pages are locked after page fault
Add mlockall flag for locking pages on fault
Add tests for lock on fault
arch/alpha/include/uapi/asm/mman.h | 2 +
arch/mips/include/uapi/asm/mman.h | 2 +
arch/parisc/include/uapi/asm/mman.h | 2 +
arch/powerpc/include/uapi/asm/mman.h | 2 +
arch/sparc/include/uapi/asm/mman.h | 2 +
arch/tile/include/uapi/asm/mman.h | 2 +
arch/xtensa/include/uapi/asm/mman.h | 2 +
include/linux/mm.h | 1 +
include/linux/mman.h | 3 +-
include/uapi/asm-generic/mman.h | 2 +
mm/mlock.c | 13 ++-
mm/mmap.c | 4 +-
mm/swap.c | 3 +-
tools/testing/selftests/vm/Makefile | 8 +-
tools/testing/selftests/vm/lock-on-fault.c | 145 ++++++++++++++++++++++++++++
tools/testing/selftests/vm/on-fault-limit.c | 47 +++++++++
tools/testing/selftests/vm/run_vmtests | 23 +++++
17 files changed, 254 insertions(+), 9 deletions(-)
create mode 100644 tools/testing/selftests/vm/lock-on-fault.c
create mode 100644 tools/testing/selftests/vm/on-fault-limit.c
Cc: Shuah Khan <shuahkh@osg.samsung.com>
Cc: linux-alpha@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Cc: linux-mips@linux-mips.org
Cc: linux-parisc@vger.kernel.org
Cc: linuxppc-dev@lists.ozlabs.org
Cc: sparclinux@vger.kernel.org
Cc: linux-xtensa@linux-xtensa.org
Cc: linux-mm@kvack.org
Cc: linux-arch@vger.kernel.org
Cc: linux-api@vger.kernel.org
--
1.9.1
^ permalink raw reply
* [PATCH 1/3] Add flag to request pages are locked after page fault
From: Eric B Munson @ 2015-05-08 19:33 UTC (permalink / raw)
To: Andrew Morton
Cc: Eric B Munson, linux-alpha, linux-kernel, linux-mips,
linux-parisc, linuxppc-dev, sparclinux, linux-xtensa, linux-mm,
linux-arch, linux-api
In-Reply-To: <1431113626-19153-1-git-send-email-emunson@akamai.com>
The cost of faulting in all memory to be locked can be very high when
working with large mappings. If only portions of the mapping will be
used this can incur a high penalty for locking. This patch introduces
the ability to request that pages are not pre-faulted, but are placed on
the unevictable LRU when they are finally faulted in.
To keep accounting checks out of the page fault path, users are billed
for the entire mapping lock as if MAP_LOCKED was used.
Signed-off-by: Eric B Munson <emunson@akamai.com>
Cc: linux-alpha@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Cc: linux-mips@linux-mips.org
Cc: linux-parisc@vger.kernel.org
Cc: linuxppc-dev@lists.ozlabs.org
Cc: sparclinux@vger.kernel.org
Cc: linux-xtensa@linux-xtensa.org
Cc: linux-mm@kvack.org
Cc: linux-arch@vger.kernel.org
Cc: linux-api@vger.kernel.org
---
arch/alpha/include/uapi/asm/mman.h | 1 +
arch/mips/include/uapi/asm/mman.h | 1 +
arch/parisc/include/uapi/asm/mman.h | 1 +
arch/powerpc/include/uapi/asm/mman.h | 1 +
arch/sparc/include/uapi/asm/mman.h | 1 +
arch/tile/include/uapi/asm/mman.h | 1 +
arch/xtensa/include/uapi/asm/mman.h | 1 +
include/linux/mm.h | 1 +
include/linux/mman.h | 3 ++-
include/uapi/asm-generic/mman.h | 1 +
mm/mmap.c | 4 ++--
mm/swap.c | 3 ++-
12 files changed, 15 insertions(+), 4 deletions(-)
diff --git a/arch/alpha/include/uapi/asm/mman.h b/arch/alpha/include/uapi/asm/mman.h
index 0086b47..15e96e1 100644
--- a/arch/alpha/include/uapi/asm/mman.h
+++ b/arch/alpha/include/uapi/asm/mman.h
@@ -30,6 +30,7 @@
#define MAP_NONBLOCK 0x40000 /* do not block on IO */
#define MAP_STACK 0x80000 /* give out an address that is best suited for process/thread stacks */
#define MAP_HUGETLB 0x100000 /* create a huge page mapping */
+#define MAP_LOCKONFAULT 0x200000 /* Lock pages after they are faulted in, do not prefault */
#define MS_ASYNC 1 /* sync memory asynchronously */
#define MS_SYNC 2 /* synchronous memory sync */
diff --git a/arch/mips/include/uapi/asm/mman.h b/arch/mips/include/uapi/asm/mman.h
index cfcb876..47846a5 100644
--- a/arch/mips/include/uapi/asm/mman.h
+++ b/arch/mips/include/uapi/asm/mman.h
@@ -48,6 +48,7 @@
#define MAP_NONBLOCK 0x20000 /* do not block on IO */
#define MAP_STACK 0x40000 /* give out an address that is best suited for process/thread stacks */
#define MAP_HUGETLB 0x80000 /* create a huge page mapping */
+#define MAP_LOCKONFAULT 0x100000 /* Lock pages after they are faulted in, do not prefault */
/*
* Flags for msync
diff --git a/arch/parisc/include/uapi/asm/mman.h b/arch/parisc/include/uapi/asm/mman.h
index 294d251..1514cd7 100644
--- a/arch/parisc/include/uapi/asm/mman.h
+++ b/arch/parisc/include/uapi/asm/mman.h
@@ -24,6 +24,7 @@
#define MAP_NONBLOCK 0x20000 /* do not block on IO */
#define MAP_STACK 0x40000 /* give out an address that is best suited for process/thread stacks */
#define MAP_HUGETLB 0x80000 /* create a huge page mapping */
+#define MAP_LOCKONFAULT 0x100000 /* Lock pages after they are faulted in, do not prefault */
#define MS_SYNC 1 /* synchronous memory sync */
#define MS_ASYNC 2 /* sync memory asynchronously */
diff --git a/arch/powerpc/include/uapi/asm/mman.h b/arch/powerpc/include/uapi/asm/mman.h
index 6ea26df..fce74fe 100644
--- a/arch/powerpc/include/uapi/asm/mman.h
+++ b/arch/powerpc/include/uapi/asm/mman.h
@@ -27,5 +27,6 @@
#define MAP_NONBLOCK 0x10000 /* do not block on IO */
#define MAP_STACK 0x20000 /* give out an address that is best suited for process/thread stacks */
#define MAP_HUGETLB 0x40000 /* create a huge page mapping */
+#define MAP_LOCKONFAULT 0x80000 /* Lock pages after they are faulted in, do not prefault */
#endif /* _UAPI_ASM_POWERPC_MMAN_H */
diff --git a/arch/sparc/include/uapi/asm/mman.h b/arch/sparc/include/uapi/asm/mman.h
index 0b14df3..12425d8 100644
--- a/arch/sparc/include/uapi/asm/mman.h
+++ b/arch/sparc/include/uapi/asm/mman.h
@@ -22,6 +22,7 @@
#define MAP_NONBLOCK 0x10000 /* do not block on IO */
#define MAP_STACK 0x20000 /* give out an address that is best suited for process/thread stacks */
#define MAP_HUGETLB 0x40000 /* create a huge page mapping */
+#define MAP_LOCKONFAULT 0x80000 /* Lock pages after they are faulted in, do not prefault */
#endif /* _UAPI__SPARC_MMAN_H__ */
diff --git a/arch/tile/include/uapi/asm/mman.h b/arch/tile/include/uapi/asm/mman.h
index 81b8fc3..ec04eaf 100644
--- a/arch/tile/include/uapi/asm/mman.h
+++ b/arch/tile/include/uapi/asm/mman.h
@@ -29,6 +29,7 @@
#define MAP_DENYWRITE 0x0800 /* ETXTBSY */
#define MAP_EXECUTABLE 0x1000 /* mark it as an executable */
#define MAP_HUGETLB 0x4000 /* create a huge page mapping */
+#define MAP_LOCKONFAULT 0x8000 /* Lock pages after they are faulted in, do not prefault */
/*
diff --git a/arch/xtensa/include/uapi/asm/mman.h b/arch/xtensa/include/uapi/asm/mman.h
index 201aec0..42d43cc 100644
--- a/arch/xtensa/include/uapi/asm/mman.h
+++ b/arch/xtensa/include/uapi/asm/mman.h
@@ -55,6 +55,7 @@
#define MAP_NONBLOCK 0x20000 /* do not block on IO */
#define MAP_STACK 0x40000 /* give out an address that is best suited for process/thread stacks */
#define MAP_HUGETLB 0x80000 /* create a huge page mapping */
+#define MAP_LOCKONFAULT 0x100000 /* Lock pages after they are faulted in, do not prefault */
#ifdef CONFIG_MMAP_ALLOW_UNINITIALIZED
# define MAP_UNINITIALIZED 0x4000000 /* For anonymous mmap, memory could be
* uninitialized */
diff --git a/include/linux/mm.h b/include/linux/mm.h
index 0755b9f..3e31457 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -126,6 +126,7 @@ extern unsigned int kobjsize(const void *objp);
#define VM_PFNMAP 0x00000400 /* Page-ranges managed without "struct page", just pure PFN */
#define VM_DENYWRITE 0x00000800 /* ETXTBSY on write attempts.. */
+#define VM_LOCKONFAULT 0x00001000 /* Lock the pages covered when they are faulted in */
#define VM_LOCKED 0x00002000
#define VM_IO 0x00004000 /* Memory mapped I/O or similar */
diff --git a/include/linux/mman.h b/include/linux/mman.h
index 16373c8..437264b 100644
--- a/include/linux/mman.h
+++ b/include/linux/mman.h
@@ -86,7 +86,8 @@ calc_vm_flag_bits(unsigned long flags)
{
return _calc_vm_trans(flags, MAP_GROWSDOWN, VM_GROWSDOWN ) |
_calc_vm_trans(flags, MAP_DENYWRITE, VM_DENYWRITE ) |
- _calc_vm_trans(flags, MAP_LOCKED, VM_LOCKED );
+ _calc_vm_trans(flags, MAP_LOCKED, VM_LOCKED ) |
+ _calc_vm_trans(flags, MAP_LOCKONFAULT,VM_LOCKONFAULT);
}
unsigned long vm_commit_limit(void);
diff --git a/include/uapi/asm-generic/mman.h b/include/uapi/asm-generic/mman.h
index e9fe6fd..fc4e586 100644
--- a/include/uapi/asm-generic/mman.h
+++ b/include/uapi/asm-generic/mman.h
@@ -12,6 +12,7 @@
#define MAP_NONBLOCK 0x10000 /* do not block on IO */
#define MAP_STACK 0x20000 /* give out an address that is best suited for process/thread stacks */
#define MAP_HUGETLB 0x40000 /* create a huge page mapping */
+#define MAP_LOCKONFAULT 0x80000 /* Lock pages after they are faulted in, do not prefault */
/* Bits [26:31] are reserved, see mman-common.h for MAP_HUGETLB usage */
diff --git a/mm/mmap.c b/mm/mmap.c
index bb50cac..ba1a6bf 100644
--- a/mm/mmap.c
+++ b/mm/mmap.c
@@ -1233,7 +1233,7 @@ static inline int mlock_future_check(struct mm_struct *mm,
unsigned long locked, lock_limit;
/* mlock MCL_FUTURE? */
- if (flags & VM_LOCKED) {
+ if (flags & (VM_LOCKED | VM_LOCKONFAULT)) {
locked = len >> PAGE_SHIFT;
locked += mm->locked_vm;
lock_limit = rlimit(RLIMIT_MEMLOCK);
@@ -1301,7 +1301,7 @@ unsigned long do_mmap_pgoff(struct file *file, unsigned long addr,
vm_flags = calc_vm_prot_bits(prot) | calc_vm_flag_bits(flags) |
mm->def_flags | VM_MAYREAD | VM_MAYWRITE | VM_MAYEXEC;
- if (flags & MAP_LOCKED)
+ if (flags & (MAP_LOCKED | MAP_LOCKONFAULT))
if (!can_do_mlock())
return -EPERM;
diff --git a/mm/swap.c b/mm/swap.c
index a7251a8..07c905e 100644
--- a/mm/swap.c
+++ b/mm/swap.c
@@ -711,7 +711,8 @@ void lru_cache_add_active_or_unevictable(struct page *page,
{
VM_BUG_ON_PAGE(PageLRU(page), page);
- if (likely((vma->vm_flags & (VM_LOCKED | VM_SPECIAL)) != VM_LOCKED)) {
+ if (likely((vma->vm_flags & (VM_LOCKED | VM_LOCKONFAULT)) == 0) ||
+ (vma->vm_flags & VM_SPECIAL)) {
SetPageActive(page);
lru_cache_add(page);
return;
--
1.9.1
^ permalink raw reply related
* [PATCH 2/3] Add mlockall flag for locking pages on fault
From: Eric B Munson @ 2015-05-08 19:33 UTC (permalink / raw)
To: Andrew Morton
Cc: Eric B Munson, linux-alpha, linux-kernel, linux-mips,
linux-parisc, linuxppc-dev, sparclinux, linux-xtensa, linux-arch,
linux-api, linux-mm
In-Reply-To: <1431113626-19153-1-git-send-email-emunson@akamai.com>
Building on the previous patch, extend mlockall() to give a process a
way to specify that pages should be locked when they are faulted in, but
that pre-faulting is not needed.
Signed-off-by: Eric B Munson <emunson@akamai.com>
Cc: linux-alpha@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Cc: linux-mips@linux-mips.org
Cc: linux-parisc@vger.kernel.org
Cc: linuxppc-dev@lists.ozlabs.org
Cc: sparclinux@vger.kernel.org
Cc: linux-xtensa@linux-xtensa.org
Cc: linux-arch@vger.kernel.org
Cc: linux-api@vger.kernel.org
Cc: linux-mm@kvack.org
---
arch/alpha/include/uapi/asm/mman.h | 1 +
arch/mips/include/uapi/asm/mman.h | 1 +
arch/parisc/include/uapi/asm/mman.h | 1 +
arch/powerpc/include/uapi/asm/mman.h | 1 +
arch/sparc/include/uapi/asm/mman.h | 1 +
arch/tile/include/uapi/asm/mman.h | 1 +
arch/xtensa/include/uapi/asm/mman.h | 1 +
include/uapi/asm-generic/mman.h | 1 +
mm/mlock.c | 13 +++++++++----
9 files changed, 17 insertions(+), 4 deletions(-)
diff --git a/arch/alpha/include/uapi/asm/mman.h b/arch/alpha/include/uapi/asm/mman.h
index 15e96e1..3120dfb 100644
--- a/arch/alpha/include/uapi/asm/mman.h
+++ b/arch/alpha/include/uapi/asm/mman.h
@@ -38,6 +38,7 @@
#define MCL_CURRENT 8192 /* lock all currently mapped pages */
#define MCL_FUTURE 16384 /* lock all additions to address space */
+#define MCL_ON_FAULT 32768 /* lock all pages that are faulted in */
#define MADV_NORMAL 0 /* no further special treatment */
#define MADV_RANDOM 1 /* expect random page references */
diff --git a/arch/mips/include/uapi/asm/mman.h b/arch/mips/include/uapi/asm/mman.h
index 47846a5..82aec3c 100644
--- a/arch/mips/include/uapi/asm/mman.h
+++ b/arch/mips/include/uapi/asm/mman.h
@@ -62,6 +62,7 @@
*/
#define MCL_CURRENT 1 /* lock all current mappings */
#define MCL_FUTURE 2 /* lock all future mappings */
+#define MCL_ON_FAULT 4 /* lock all pages that are faulted in */
#define MADV_NORMAL 0 /* no further special treatment */
#define MADV_RANDOM 1 /* expect random page references */
diff --git a/arch/parisc/include/uapi/asm/mman.h b/arch/parisc/include/uapi/asm/mman.h
index 1514cd7..f4601f3 100644
--- a/arch/parisc/include/uapi/asm/mman.h
+++ b/arch/parisc/include/uapi/asm/mman.h
@@ -32,6 +32,7 @@
#define MCL_CURRENT 1 /* lock all current mappings */
#define MCL_FUTURE 2 /* lock all future mappings */
+#define MCL_ON_FAULT 4 /* lock all pages that are faulted in */
#define MADV_NORMAL 0 /* no further special treatment */
#define MADV_RANDOM 1 /* expect random page references */
diff --git a/arch/powerpc/include/uapi/asm/mman.h b/arch/powerpc/include/uapi/asm/mman.h
index fce74fe..0a28efc 100644
--- a/arch/powerpc/include/uapi/asm/mman.h
+++ b/arch/powerpc/include/uapi/asm/mman.h
@@ -22,6 +22,7 @@
#define MCL_CURRENT 0x2000 /* lock all currently mapped pages */
#define MCL_FUTURE 0x4000 /* lock all additions to address space */
+#define MCL_ON_FAULT 0x80000 /* lock all pages that are faulted in */
#define MAP_POPULATE 0x8000 /* populate (prefault) pagetables */
#define MAP_NONBLOCK 0x10000 /* do not block on IO */
diff --git a/arch/sparc/include/uapi/asm/mman.h b/arch/sparc/include/uapi/asm/mman.h
index 12425d8..119be80 100644
--- a/arch/sparc/include/uapi/asm/mman.h
+++ b/arch/sparc/include/uapi/asm/mman.h
@@ -17,6 +17,7 @@
#define MCL_CURRENT 0x2000 /* lock all currently mapped pages */
#define MCL_FUTURE 0x4000 /* lock all additions to address space */
+#define MCL_ON_FAULT 0x80000 /* lock all pages that are faulted in */
#define MAP_POPULATE 0x8000 /* populate (prefault) pagetables */
#define MAP_NONBLOCK 0x10000 /* do not block on IO */
diff --git a/arch/tile/include/uapi/asm/mman.h b/arch/tile/include/uapi/asm/mman.h
index ec04eaf..66ea935 100644
--- a/arch/tile/include/uapi/asm/mman.h
+++ b/arch/tile/include/uapi/asm/mman.h
@@ -37,6 +37,7 @@
*/
#define MCL_CURRENT 1 /* lock all current mappings */
#define MCL_FUTURE 2 /* lock all future mappings */
+#define MCL_ON_FAULT 4 /* lock all pages that are faulted in */
#endif /* _ASM_TILE_MMAN_H */
diff --git a/arch/xtensa/include/uapi/asm/mman.h b/arch/xtensa/include/uapi/asm/mman.h
index 42d43cc..9abcc29 100644
--- a/arch/xtensa/include/uapi/asm/mman.h
+++ b/arch/xtensa/include/uapi/asm/mman.h
@@ -75,6 +75,7 @@
*/
#define MCL_CURRENT 1 /* lock all current mappings */
#define MCL_FUTURE 2 /* lock all future mappings */
+#define MCL_ON_FAULT 4 /* lock all pages that are faulted in */
#define MADV_NORMAL 0 /* no further special treatment */
#define MADV_RANDOM 1 /* expect random page references */
diff --git a/include/uapi/asm-generic/mman.h b/include/uapi/asm-generic/mman.h
index fc4e586..6ac7a7b 100644
--- a/include/uapi/asm-generic/mman.h
+++ b/include/uapi/asm-generic/mman.h
@@ -18,5 +18,6 @@
#define MCL_CURRENT 1 /* lock all current mappings */
#define MCL_FUTURE 2 /* lock all future mappings */
+#define MCL_ON_FAULT 4 /* lock all pages that are faulted in */
#endif /* __ASM_GENERIC_MMAN_H */
diff --git a/mm/mlock.c b/mm/mlock.c
index 6fd2cf1..1406835 100644
--- a/mm/mlock.c
+++ b/mm/mlock.c
@@ -579,7 +579,7 @@ static int do_mlock(unsigned long start, size_t len, int on)
/* Here we know that vma->vm_start <= nstart < vma->vm_end. */
- newflags = vma->vm_flags & ~VM_LOCKED;
+ newflags = vma->vm_flags & ~(VM_LOCKED | VM_LOCKONFAULT);
if (on)
newflags |= VM_LOCKED;
@@ -662,13 +662,17 @@ static int do_mlockall(int flags)
current->mm->def_flags |= VM_LOCKED;
else
current->mm->def_flags &= ~VM_LOCKED;
- if (flags == MCL_FUTURE)
+ if (flags & MCL_ON_FAULT)
+ current->mm->def_flags |= VM_LOCKONFAULT;
+ else
+ current->mm->def_flags &= ~VM_LOCKONFAULT;
+ if (flags == MCL_FUTURE || flags == MCL_ON_FAULT)
goto out;
for (vma = current->mm->mmap; vma ; vma = prev->vm_next) {
vm_flags_t newflags;
- newflags = vma->vm_flags & ~VM_LOCKED;
+ newflags = vma->vm_flags & ~(VM_LOCKED | VM_LOCKONFAULT);
if (flags & MCL_CURRENT)
newflags |= VM_LOCKED;
@@ -685,7 +689,8 @@ SYSCALL_DEFINE1(mlockall, int, flags)
unsigned long lock_limit;
int ret = -EINVAL;
- if (!flags || (flags & ~(MCL_CURRENT | MCL_FUTURE)))
+ if (!flags || (flags & ~(MCL_CURRENT | MCL_FUTURE | MCL_ON_FAULT)) ||
+ ((flags & MCL_FUTURE) && (flags & MCL_ON_FAULT)))
goto out;
ret = -EPERM;
--
1.9.1
^ permalink raw reply related
* [PATCH 3/3] Add tests for lock on fault
From: Eric B Munson @ 2015-05-08 19:33 UTC (permalink / raw)
To: Andrew Morton
Cc: Eric B Munson, Shuah Khan, linux-mm-Bw31MaZKKs3YtjvyW6yDsg,
linux-kernel-u79uwXL29TY76Z2rM5mHXA,
linux-api-u79uwXL29TY76Z2rM5mHXA
In-Reply-To: <1431113626-19153-1-git-send-email-emunson-JqFfY2XvxFXQT0dZR+AlfA@public.gmane.org>
Test the mmap() flag, the mlockall() flag, and ensure that mlock limits
are respected. Note that the limit test needs to be run a normal user.
Signed-off-by: Eric B Munson <emunson-JqFfY2XvxFXQT0dZR+AlfA@public.gmane.org>
Cc: Shuah Khan <shuahkh-JPH+aEBZ4P+UEJcrhfAQsw@public.gmane.org>
Cc: linux-mm-Bw31MaZKKs3YtjvyW6yDsg@public.gmane.org
Cc: linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
Cc: linux-api-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
---
tools/testing/selftests/vm/Makefile | 8 +-
tools/testing/selftests/vm/lock-on-fault.c | 145 ++++++++++++++++++++++++++++
tools/testing/selftests/vm/on-fault-limit.c | 47 +++++++++
tools/testing/selftests/vm/run_vmtests | 23 +++++
4 files changed, 222 insertions(+), 1 deletion(-)
create mode 100644 tools/testing/selftests/vm/lock-on-fault.c
create mode 100644 tools/testing/selftests/vm/on-fault-limit.c
diff --git a/tools/testing/selftests/vm/Makefile b/tools/testing/selftests/vm/Makefile
index a5ce953..32f3d20 100644
--- a/tools/testing/selftests/vm/Makefile
+++ b/tools/testing/selftests/vm/Makefile
@@ -1,7 +1,13 @@
# Makefile for vm selftests
CFLAGS = -Wall
-BINARIES = hugepage-mmap hugepage-shm map_hugetlb thuge-gen hugetlbfstest
+BINARIES = hugepage-mmap
+BINARIES += hugepage-shm
+BINARIES += hugetlbfstest
+BINARIES += lock-on-fault
+BINARIES += map_hugetlb
+BINARIES += on-fault-limit
+BINARIES += thuge-gen
BINARIES += transhuge-stress
all: $(BINARIES)
diff --git a/tools/testing/selftests/vm/lock-on-fault.c b/tools/testing/selftests/vm/lock-on-fault.c
new file mode 100644
index 0000000..e6a9688
--- /dev/null
+++ b/tools/testing/selftests/vm/lock-on-fault.c
@@ -0,0 +1,145 @@
+#include <sys/mman.h>
+#include <stdio.h>
+#include <unistd.h>
+#include <string.h>
+#include <sys/time.h>
+#include <sys/resource.h>
+
+#ifndef MCL_ON_FAULT
+#define MCL_ON_FAULT 4
+#endif
+
+#define PRESENT_BIT 0x8000000000000000
+#define PFN_MASK 0x007FFFFFFFFFFFFF
+#define UNEVICTABLE_BIT (1UL << 18)
+
+static int check_pageflags(void *map)
+{
+ FILE *file;
+ unsigned long pfn1;
+ unsigned long pfn2;
+ unsigned long offset1;
+ unsigned long offset2;
+ int ret = 1;
+
+ file = fopen("/proc/self/pagemap", "r");
+ if (!file) {
+ perror("fopen");
+ return ret;
+ }
+ offset1 = (unsigned long)map / getpagesize() * sizeof(unsigned long);
+ offset2 = ((unsigned long)map + getpagesize()) / getpagesize() * sizeof(unsigned long);
+ if (fseek(file, offset1, SEEK_SET)) {
+ perror("fseek");
+ goto out;
+ }
+
+ if (fread(&pfn1, sizeof(unsigned long), 1, file) != 1) {
+ perror("fread");
+ goto out;
+ }
+
+ if (fseek(file, offset2, SEEK_SET)) {
+ perror("fseek");
+ goto out;
+ }
+
+ if (fread(&pfn2, sizeof(unsigned long), 1, file) != 1) {
+ perror("fread");
+ goto out;
+ }
+
+ /* pfn2 should not be present */
+ if (pfn2 & PRESENT_BIT) {
+ printf("page map says 0x%lx\n", pfn2);
+ printf("present is 0x%lx\n", PRESENT_BIT);
+ goto out;
+ }
+
+ /* pfn1 should be present */
+ if ((pfn1 & PRESENT_BIT) == 0) {
+ printf("page map says 0x%lx\n", pfn1);
+ printf("present is 0x%lx\n", PRESENT_BIT);
+ goto out;
+ }
+
+ pfn1 &= PFN_MASK;
+ fclose(file);
+ file = fopen("/proc/kpageflags", "r");
+ if (!file) {
+ perror("fopen");
+ munmap(map, 2 * getpagesize());
+ return ret;
+ }
+
+ if (fseek(file, pfn1 * sizeof(unsigned long), SEEK_SET)) {
+ perror("fseek");
+ goto out;
+ }
+
+ if (fread(&pfn2, sizeof(unsigned long), 1, file) != 1) {
+ perror("fread");
+ goto out;
+ }
+
+ /* pfn2 now contains the entry from kpageflags for the first page, the
+ * unevictable bit should be set */
+ if ((pfn2 & UNEVICTABLE_BIT) == 0) {
+ printf("kpageflags says 0x%lx\n", pfn2);
+ printf("unevictable is 0x%lx\n", UNEVICTABLE_BIT);
+ goto out;
+ }
+
+ ret = 0;
+
+out:
+ fclose(file);
+ return ret;
+}
+
+static int test_mmap(int flags)
+{
+ int ret = 1;
+ void *map;
+
+ map = mmap(NULL, 2 * getpagesize(), PROT_READ | PROT_WRITE, flags, 0, 0);
+ if (map == MAP_FAILED) {
+ perror("mmap()");
+ return ret;
+ }
+
+ /* Write something into the first page to ensure it is present */
+ *(char *)map = 1;
+
+ ret = check_pageflags(map);
+
+ munmap(map, 2 * getpagesize());
+ return ret;
+}
+
+static int test_mlockall(void)
+{
+ int ret = 1;
+
+ if (mlockall(MCL_ON_FAULT)) {
+ perror("mlockall");
+ return ret;
+ }
+
+ ret = test_mmap(MAP_PRIVATE | MAP_ANONYMOUS);
+ munlockall();
+ return ret;
+}
+
+#ifndef MAP_LOCKONFAULT
+#define MAP_LOCKONFAULT (MAP_HUGETLB << 1)
+#endif
+
+int main(int argc, char **argv)
+{
+ int ret = 0;
+
+ ret += test_mmap(MAP_PRIVATE | MAP_ANONYMOUS | MAP_LOCKONFAULT);
+ ret += test_mlockall();
+ return ret;
+}
diff --git a/tools/testing/selftests/vm/on-fault-limit.c b/tools/testing/selftests/vm/on-fault-limit.c
new file mode 100644
index 0000000..bd70078
--- /dev/null
+++ b/tools/testing/selftests/vm/on-fault-limit.c
@@ -0,0 +1,47 @@
+#include <sys/mman.h>
+#include <stdio.h>
+#include <unistd.h>
+#include <string.h>
+#include <sys/time.h>
+#include <sys/resource.h>
+
+#ifndef MCL_ON_FAULT
+#define MCL_ON_FAULT 4
+#endif
+
+static int test_limit(void)
+{
+ int ret = 1;
+ struct rlimit lims;
+ void *map;
+
+ if (getrlimit(RLIMIT_MEMLOCK, &lims)) {
+ perror("getrlimit");
+ return ret;
+ }
+
+ if (mlockall(MCL_ON_FAULT)) {
+ perror("mlockall");
+ return ret;
+ }
+
+ map = mmap(NULL, 2 * lims.rlim_max, PROT_READ | PROT_WRITE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_POPULATE, 0, 0);
+ if (map != MAP_FAILED)
+ printf("mmap should have failed, but didn't\n");
+ else {
+ ret = 0;
+ munmap(map, 2 * lims.rlim_max);
+ }
+
+ munlockall();
+ return ret;
+}
+
+int main(int argc, char **argv)
+{
+ int ret = 0;
+
+ ret += test_limit();
+ return ret;
+}
diff --git a/tools/testing/selftests/vm/run_vmtests b/tools/testing/selftests/vm/run_vmtests
index c87b681..c1aecce 100755
--- a/tools/testing/selftests/vm/run_vmtests
+++ b/tools/testing/selftests/vm/run_vmtests
@@ -90,4 +90,27 @@ fi
umount $mnt
rm -rf $mnt
echo $nr_hugepgs > /proc/sys/vm/nr_hugepages
+
+echo "--------------------"
+echo "running lock-on-fault"
+echo "--------------------"
+./lock-on-fault
+if [ $? -ne 0 ]; then
+ echo "[FAIL]"
+ exitcode=1
+else
+ echo "[PASS]"
+fi
+
+echo "--------------------"
+echo "running on-fault-limit"
+echo "--------------------"
+sudo -u nobody ./on-fault-limit
+if [ $? -ne 0 ]; then
+ echo "[FAIL]"
+ exitcode=1
+else
+ echo "[PASS]"
+fi
+
exit $exitcode
--
1.9.1
^ permalink raw reply related
* Re: [PATCH 0/3] Allow user to request memory to be locked on page fault
From: Andrew Morton @ 2015-05-08 19:42 UTC (permalink / raw)
To: Eric B Munson
Cc: Shuah Khan, linux-alpha, linux-kernel, linux-mips, linux-parisc,
linuxppc-dev, sparclinux, linux-xtensa, linux-mm, linux-arch,
linux-api
In-Reply-To: <1431113626-19153-1-git-send-email-emunson@akamai.com>
On Fri, 8 May 2015 15:33:43 -0400 Eric B Munson <emunson@akamai.com> wrote:
> mlock() allows a user to control page out of program memory, but this
> comes at the cost of faulting in the entire mapping when it is
> allocated. For large mappings where the entire area is not necessary
> this is not ideal.
>
> This series introduces new flags for mmap() and mlockall() that allow a
> user to specify that the covered are should not be paged out, but only
> after the memory has been used the first time.
Please tell us much much more about the value of these changes: the use
cases, the behavioural improvements and performance results which the
patchset brings to those use cases, etc.
^ permalink raw reply
* Re: [PATCH 0/3] Allow user to request memory to be locked on page fault
From: Eric B Munson @ 2015-05-08 20:06 UTC (permalink / raw)
To: Andrew Morton
Cc: Shuah Khan, linux-alpha-u79uwXL29TY76Z2rM5mHXA,
linux-kernel-u79uwXL29TY76Z2rM5mHXA,
linux-mips-6z/3iImG2C8G8FEW9MqTrA,
linux-parisc-u79uwXL29TY76Z2rM5mHXA,
linuxppc-dev-uLR06cmDAlY/bJ5BZ2RsiQ,
sparclinux-u79uwXL29TY76Z2rM5mHXA,
linux-xtensa-PjhNF2WwrV/0Sa2dR60CXw,
linux-mm-Bw31MaZKKs3YtjvyW6yDsg,
linux-arch-u79uwXL29TY76Z2rM5mHXA,
linux-api-u79uwXL29TY76Z2rM5mHXA
In-Reply-To: <20150508124203.6679b1d35ad9555425003929-de/tnXTf+JLsfHDXvbKv3WD2FQJk+8+b@public.gmane.org>
[-- Attachment #1: Type: text/plain, Size: 1552 bytes --]
On Fri, 08 May 2015, Andrew Morton wrote:
> On Fri, 8 May 2015 15:33:43 -0400 Eric B Munson <emunson-JqFfY2XvxFXQT0dZR+AlfA@public.gmane.org> wrote:
>
> > mlock() allows a user to control page out of program memory, but this
> > comes at the cost of faulting in the entire mapping when it is
> > allocated. For large mappings where the entire area is not necessary
> > this is not ideal.
> >
> > This series introduces new flags for mmap() and mlockall() that allow a
> > user to specify that the covered are should not be paged out, but only
> > after the memory has been used the first time.
>
> Please tell us much much more about the value of these changes: the use
> cases, the behavioural improvements and performance results which the
> patchset brings to those use cases, etc.
>
The primary use case is for mmaping large files read only. The process
knows that some of the data is necessary, but it is unlikely that the
entire file will be needed. The developer only wants to pay the cost to
read the data in once. Unfortunately developer must choose between
allowing the kernel to page in the memory as needed and guaranteeing
that the data will only be read from disk once. The first option runs
the risk of having the memory reclaimed if the system is under memory
pressure, the second forces the memory usage and startup delay when
faulting in the entire file.
I am working on getting startup times with and without this change for
an application, I will post them as soon as I have them.
Eric
[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]
^ permalink raw reply
* Re: [PATCH 0/3] Allow user to request memory to be locked on page fault
From: Andrew Morton @ 2015-05-08 20:15 UTC (permalink / raw)
To: Eric B Munson
Cc: Shuah Khan, linux-alpha, linux-kernel, linux-mips, linux-parisc,
linuxppc-dev, sparclinux, linux-xtensa, linux-mm, linux-arch,
linux-api
In-Reply-To: <20150508200610.GB29933@akamai.com>
On Fri, 8 May 2015 16:06:10 -0400 Eric B Munson <emunson@akamai.com> wrote:
> On Fri, 08 May 2015, Andrew Morton wrote:
>
> > On Fri, 8 May 2015 15:33:43 -0400 Eric B Munson <emunson@akamai.com> wrote:
> >
> > > mlock() allows a user to control page out of program memory, but this
> > > comes at the cost of faulting in the entire mapping when it is
> > > allocated. For large mappings where the entire area is not necessary
> > > this is not ideal.
> > >
> > > This series introduces new flags for mmap() and mlockall() that allow a
> > > user to specify that the covered are should not be paged out, but only
> > > after the memory has been used the first time.
> >
> > Please tell us much much more about the value of these changes: the use
> > cases, the behavioural improvements and performance results which the
> > patchset brings to those use cases, etc.
> >
>
> The primary use case is for mmaping large files read only. The process
> knows that some of the data is necessary, but it is unlikely that the
> entire file will be needed. The developer only wants to pay the cost to
> read the data in once. Unfortunately developer must choose between
> allowing the kernel to page in the memory as needed and guaranteeing
> that the data will only be read from disk once. The first option runs
> the risk of having the memory reclaimed if the system is under memory
> pressure, the second forces the memory usage and startup delay when
> faulting in the entire file.
Why can't the application mmap only those parts of the file which it
wants and mlock those?
> I am working on getting startup times with and without this change for
> an application, I will post them as soon as I have them.
^ permalink raw reply
* Re: [PATCH 0/6] support "dataplane" mode for nohz_full
From: Andrew Morton @ 2015-05-08 21:18 UTC (permalink / raw)
To: Chris Metcalf
Cc: Gilad Ben Yossef, Steven Rostedt, Ingo Molnar, Peter Zijlstra,
Rik van Riel, Tejun Heo, Frederic Weisbecker, Thomas Gleixner,
Paul E. McKenney, Christoph Lameter, Srivatsa S. Bhat,
linux-doc-u79uwXL29TY76Z2rM5mHXA,
linux-api-u79uwXL29TY76Z2rM5mHXA,
linux-kernel-u79uwXL29TY76Z2rM5mHXA
In-Reply-To: <1431107927-13998-1-git-send-email-cmetcalf-d5a29ZRxExrQT0dZR+AlfA@public.gmane.org>
On Fri, 8 May 2015 13:58:41 -0400 Chris Metcalf <cmetcalf-d5a29ZRxExrQT0dZR+AlfA@public.gmane.org> wrote:
> A prctl() option (PR_SET_DATAPLANE) is added
Dumb question: what does the term "dataplane" mean in this context? I
can't see the relationship between those words and what this patch
does.
^ permalink raw reply
* Re: [PATCH 0/6] support "dataplane" mode for nohz_full
From: Steven Rostedt @ 2015-05-08 21:22 UTC (permalink / raw)
To: Andrew Morton
Cc: Chris Metcalf, Gilad Ben Yossef, Ingo Molnar, Peter Zijlstra,
Rik van Riel, Tejun Heo, Frederic Weisbecker, Thomas Gleixner,
Paul E. McKenney, Christoph Lameter, Srivatsa S. Bhat, linux-doc,
linux-api, linux-kernel
In-Reply-To: <20150508141824.797eb0d89d514e39fd30fffe@linux-foundation.org>
On Fri, 8 May 2015 14:18:24 -0700
Andrew Morton <akpm@linux-foundation.org> wrote:
> On Fri, 8 May 2015 13:58:41 -0400 Chris Metcalf <cmetcalf@ezchip.com> wrote:
>
> > A prctl() option (PR_SET_DATAPLANE) is added
>
> Dumb question: what does the term "dataplane" mean in this context? I
> can't see the relationship between those words and what this patch
> does.
I was thinking the same thing. I haven't gotten around to searching
DATAPLANE yet.
I would assume we want a name that is more meaningful for what is
happening.
-- Steve
^ permalink raw reply
* [GIT PULL v2] selftest: Add futex functional tests
From: Darren Hart @ 2015-05-08 22:09 UTC (permalink / raw)
To: linux-api-u79uwXL29TY76Z2rM5mHXA, Linux Kernel Mailing List
Cc: Shuah Khan, Ingo Molnar, Peter Zijlstra, Thomas Gleixner,
Davidlohr Bueso, KOSAKI Motohiro, Darren Hart
Hi Shuah,
This series begins the process of migrating my futextest tests into kselftest.
I've started with only the functional tests, as the performance and stress may
not be appropriate for kselftest as they stand.
I cleaned up various complaints from checkpatch, but I ignored others that would
require significant rework of the testcases, such as not using volatile and not
creating new typedefs.
Since v1:
Avoid checkpatch errors on 1/5 by:
- combining a later patch which did substantial cleanup.
- removing file-local typedefs and replacing with structs
- correcting all >80 char lines, except for quoted strings and header boiler
plate due to long email addresses
I did *not* make changes for the following:
- Use of new typdefs for types futex_t and atomic_t as they are used throughout
the test suite and I consider them to be worth while.
- Use of volatile as the warning is about use of volatile in kernel code. The
usage in futextest is correct, as an indicator that other threads may modify
the value.
- Adding parentheses around complex defines as it would break one use case and
change the behavior of another.
The patches will follow, but I'm providing a pull request for your convenience
as well.
The following changes since commit b787f68c36d49bb1d9236f403813641efa74a031:
Linux 4.1-rc1 (2015-04-26 17:59:10 -0700)
are available in the git repository at:
git://git.infradead.org/users/dvhart/linux.git futextest-v2
Darren Hart (5):
selftests: Add futex functional tests
selftest/futex: Update Makefile to use lib.mk
selftest/futex: Increment ksft pass and fail counters
selftest: Add futex tests to the top-level Makefile
kselftest: Add exit code defines
tools/testing/selftests/Makefile | 1 +
tools/testing/selftests/futex/Makefile | 29 ++
tools/testing/selftests/futex/README | 62 ++++
tools/testing/selftests/futex/functional/Makefile | 25 ++
.../selftests/futex/functional/futex_requeue_pi.c | 409 +++++++++++++++++++++
.../functional/futex_requeue_pi_mismatched_ops.c | 135 +++++++
.../functional/futex_requeue_pi_signal_restart.c | 223 +++++++++++
.../functional/futex_wait_private_mapped_file.c | 125 +++++++
.../futex/functional/futex_wait_timeout.c | 86 +++++
.../functional/futex_wait_uninitialized_heap.c | 124 +++++++
.../futex/functional/futex_wait_wouldblock.c | 79 ++++
tools/testing/selftests/futex/functional/run.sh | 80 ++++
tools/testing/selftests/futex/include/atomic.h | 83 +++++
tools/testing/selftests/futex/include/futextest.h | 266 ++++++++++++++
tools/testing/selftests/futex/include/logging.h | 153 ++++++++
tools/testing/selftests/futex/run.sh | 33 ++
tools/testing/selftests/kselftest.h | 17 +-
17 files changed, 1925 insertions(+), 5 deletions(-)
create mode 100644 tools/testing/selftests/futex/Makefile
create mode 100644 tools/testing/selftests/futex/README
create mode 100644 tools/testing/selftests/futex/functional/Makefile
create mode 100644 tools/testing/selftests/futex/functional/futex_requeue_pi.c
create mode 100644 tools/testing/selftests/futex/functional/futex_requeue_pi_mismatched_ops.c
create mode 100644 tools/testing/selftests/futex/functional/futex_requeue_pi_signal_restart.c
create mode 100644 tools/testing/selftests/futex/functional/futex_wait_private_mapped_file.c
create mode 100644 tools/testing/selftests/futex/functional/futex_wait_timeout.c
create mode 100644 tools/testing/selftests/futex/functional/futex_wait_uninitialized_heap.c
create mode 100644 tools/testing/selftests/futex/functional/futex_wait_wouldblock.c
create mode 100755 tools/testing/selftests/futex/functional/run.sh
create mode 100644 tools/testing/selftests/futex/include/atomic.h
create mode 100644 tools/testing/selftests/futex/include/futextest.h
create mode 100644 tools/testing/selftests/futex/include/logging.h
create mode 100755 tools/testing/selftests/futex/run.sh
--
2.1.4
^ permalink raw reply
* [PATCH 1/5] selftests: Add futex functional tests
From: Darren Hart @ 2015-05-08 22:09 UTC (permalink / raw)
To: linux-api, Linux Kernel Mailing List
Cc: Darren Hart, Shuah Khan, Ingo Molnar, Peter Zijlstra,
Thomas Gleixner, Davidlohr Bueso, KOSAKI Motohiro
In-Reply-To: <cover.1431121818.git.dvhart@linux.intel.com>
[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain; charset=y, Size: 60321 bytes --]
The futextest testsuite [1] provides functional, stress, and
performance tests for the various futex op codes. Those tests will be of
more use to futex developers if they are included with the kernel
source.
Copy the core infrastructure and the functional tests into selftests,
but adapt them for inclusion in the kernel:
- Update the Makefile to include the run_tests target, remove reference
to the performance and stress tests from the contributed sources.
- Replace my dead IBM email address with my current Intel email address.
- Remove the warrantee and write-to paragraphs from the license blurbs.
- Remove the NAME section as the filename is easily determined. ;-)
- Make the whitespace usage consistent in a couple of places.
- Cleanup various CodingStyle violations.
A future effort will explore moving the performance and stress tests
into the kernel.
1. http://git.kernel.org/cgit/linux/kernel/git/dvhart/futextest.git
Cc: Shuah Khan <shuahkh@osg.samsung.com>
Cc: linux-api@vger.kernel.org
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Davidlohr Bueso <dave@stgolabs.net>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
---
tools/testing/selftests/futex/Makefile | 11 +
tools/testing/selftests/futex/README | 62 ++++
tools/testing/selftests/futex/functional/Makefile | 24 ++
.../selftests/futex/functional/futex_requeue_pi.c | 409 +++++++++++++++++++++
.../functional/futex_requeue_pi_mismatched_ops.c | 135 +++++++
.../functional/futex_requeue_pi_signal_restart.c | 223 +++++++++++
.../functional/futex_wait_private_mapped_file.c | 125 +++++++
.../futex/functional/futex_wait_timeout.c | 86 +++++
.../functional/futex_wait_uninitialized_heap.c | 124 +++++++
.../futex/functional/futex_wait_wouldblock.c | 79 ++++
tools/testing/selftests/futex/functional/run.sh | 80 ++++
tools/testing/selftests/futex/include/atomic.h | 83 +++++
tools/testing/selftests/futex/include/futextest.h | 266 ++++++++++++++
tools/testing/selftests/futex/include/logging.h | 150 ++++++++
tools/testing/selftests/futex/run.sh | 33 ++
15 files changed, 1890 insertions(+)
create mode 100644 tools/testing/selftests/futex/Makefile
create mode 100644 tools/testing/selftests/futex/README
create mode 100644 tools/testing/selftests/futex/functional/Makefile
create mode 100644 tools/testing/selftests/futex/functional/futex_requeue_pi.c
create mode 100644 tools/testing/selftests/futex/functional/futex_requeue_pi_mismatched_ops.c
create mode 100644 tools/testing/selftests/futex/functional/futex_requeue_pi_signal_restart.c
create mode 100644 tools/testing/selftests/futex/functional/futex_wait_private_mapped_file.c
create mode 100644 tools/testing/selftests/futex/functional/futex_wait_timeout.c
create mode 100644 tools/testing/selftests/futex/functional/futex_wait_uninitialized_heap.c
create mode 100644 tools/testing/selftests/futex/functional/futex_wait_wouldblock.c
create mode 100755 tools/testing/selftests/futex/functional/run.sh
create mode 100644 tools/testing/selftests/futex/include/atomic.h
create mode 100644 tools/testing/selftests/futex/include/futextest.h
create mode 100644 tools/testing/selftests/futex/include/logging.h
create mode 100755 tools/testing/selftests/futex/run.sh
diff --git a/tools/testing/selftests/futex/Makefile b/tools/testing/selftests/futex/Makefile
new file mode 100644
index 0000000..2c26d59
--- /dev/null
+++ b/tools/testing/selftests/futex/Makefile
@@ -0,0 +1,11 @@
+SUBDIRS := functional
+
+.PHONY: all clean
+all:
+ for DIR in $(SUBDIRS); do $(MAKE) -C $$DIR $@ ; done
+
+run_tests: all
+ ./run.sh
+
+clean:
+ for DIR in $(SUBDIRS); do $(MAKE) -C $$DIR $@ ; done
diff --git a/tools/testing/selftests/futex/README b/tools/testing/selftests/futex/README
new file mode 100644
index 0000000..3224a04
--- /dev/null
+++ b/tools/testing/selftests/futex/README
@@ -0,0 +1,62 @@
+Futex Test
+==========
+Futex Test is intended to thoroughly test the Linux kernel futex system call
+API.
+
+Functional tests shall test the documented behavior of the futex operation
+code under test. This includes checking for proper behavior under normal use,
+odd corner cases, regression tests, and abject abuse and misuse.
+
+Futextest will also provide example implementation of mutual exclusion
+primitives. These can be used as is in user applications or can serve as
+examples for system libraries. These will likely be added to either a new lib/
+directory or purely as header files under include/, I'm leaning toward the
+latter.
+
+Quick Start
+-----------
+# make
+# ./run.sh
+
+Design and Implementation Goals
+-------------------------------
+o Tests should be as self contained as is practical so as to facilitate sharing
+ the individual tests on mailing list discussions and bug reports.
+o The build system shall remain as simple as possible, avoiding any archive or
+ shared object building and linking.
+o Where possible, any helper functions or other package-wide code shall be
+ implemented in header files, avoiding the need to compile intermediate object
+ files.
+o External dependendencies shall remain as minimal as possible. Currently gcc
+ and glibc are the only dependencies.
+o Tests return 0 for success and < 0 for failure.
+
+Output Formatting
+-----------------
+Test output shall be easily parsable by both human and machine. Title and
+results are printed to stdout, while intermediate ERROR or FAIL messages are
+sent to stderr. Tests shall support the -c option to print PASS, FAIL, and
+ERROR strings in color for easy visual parsing. Output shall conform to the
+following format:
+
+test_name: Description of the test
+ Arguments: arg1=val1 #units specified for clarity where appropriate
+ ERROR: Description of unexpected error
+ FAIL: Reason for test failure
+ # FIXME: Perhaps an " INFO: informational message" option would be
+ # useful here. Using -v to toggle it them on and off, as with -c.
+ # there may be multiple ERROR or FAIL messages
+Result: (PASS|FAIL|ERROR)
+
+Naming
+------
+o FIXME: decide on a sane test naming scheme. Currently the tests are named
+ based on the primary futex operation they test. Eventually this will become a
+ problem as we intend to write multiple tests which collide in this namespace.
+ Perhaps something like "wait-wake-1" "wait-wake-2" is adequate, leaving the
+ detailed description in the test source and the output.
+
+Coding Style
+------------
+o The Futex Test project adheres to the coding standards set forth by Linux
+ kernel as defined in the Linux source Documentation/CodingStyle.
diff --git a/tools/testing/selftests/futex/functional/Makefile b/tools/testing/selftests/futex/functional/Makefile
new file mode 100644
index 0000000..4098340
--- /dev/null
+++ b/tools/testing/selftests/futex/functional/Makefile
@@ -0,0 +1,24 @@
+INCLUDES := -I../include
+CFLAGS := $(CFLAGS) -g -O2 -Wall -D_GNU_SOURCE $(INCLUDES)
+LDFLAGS := $(LDFLAGS) -lpthread -lrt
+
+HEADERS := ../include/futextest.h
+TARGETS := \
+ futex_wait_timeout \
+ futex_wait_wouldblock \
+ futex_requeue_pi \
+ futex_requeue_pi_signal_restart \
+ futex_requeue_pi_mismatched_ops \
+ futex_wait_uninitialized_heap \
+ futex_wait_private_mapped_file
+
+.PHONY: all clean
+all: $(TARGETS)
+
+$(TARGETS): $(HEADERS)
+
+run_tests: all
+ ./run.sh
+
+clean:
+ rm -f $(TARGETS)
diff --git a/tools/testing/selftests/futex/functional/futex_requeue_pi.c b/tools/testing/selftests/futex/functional/futex_requeue_pi.c
new file mode 100644
index 0000000..3da06ad
--- /dev/null
+++ b/tools/testing/selftests/futex/functional/futex_requeue_pi.c
@@ -0,0 +1,409 @@
+/******************************************************************************
+ *
+ * Copyright © International Business Machines Corp., 2006-2008
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * DESCRIPTION
+ * This test excercises the futex syscall op codes needed for requeuing
+ * priority inheritance aware POSIX condition variables and mutexes.
+ *
+ * AUTHORS
+ * Sripathi Kodi <sripathik@in.ibm.com>
+ * Darren Hart <dvhart@linux.intel.com>
+ *
+ * HISTORY
+ * 2008-Jan-13: Initial version by Sripathi Kodi <sripathik@in.ibm.com>
+ * 2009-Nov-6: futex test adaptation by Darren Hart <dvhart@linux.intel.com>
+ *
+ *****************************************************************************/
+
+#include <errno.h>
+#include <limits.h>
+#include <pthread.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <signal.h>
+#include <string.h>
+#include "atomic.h"
+#include "futextest.h"
+#include "logging.h"
+
+#define MAX_WAKE_ITERS 1000
+#define THREAD_MAX 10
+#define SIGNAL_PERIOD_US 100
+
+atomic_t waiters_blocked = ATOMIC_INITIALIZER;
+atomic_t waiters_woken = ATOMIC_INITIALIZER;
+
+futex_t f1 = FUTEX_INITIALIZER;
+futex_t f2 = FUTEX_INITIALIZER;
+futex_t wake_complete = FUTEX_INITIALIZER;
+
+/* Test option defaults */
+static long timeout_ns;
+static int broadcast;
+static int owner;
+static int locked;
+
+struct thread_arg {
+ long id;
+ struct timespec *timeout;
+ int lock;
+ int ret;
+};
+#define THREAD_ARG_INITIALIZER { 0, NULL, 0, 0 }
+
+void usage(char *prog)
+{
+ printf("Usage: %s\n", prog);
+ printf(" -b Broadcast wakeup (all waiters)\n");
+ printf(" -c Use color\n");
+ printf(" -h Display this help message\n");
+ printf(" -l Lock the pi futex across requeue\n");
+ printf(" -o Use a third party pi futex owner during requeue (cancels -l)\n");
+ printf(" -t N Timeout in nanoseconds (default: 0)\n");
+ printf(" -v L Verbosity level: %d=QUIET %d=CRITICAL %d=INFO\n",
+ VQUIET, VCRITICAL, VINFO);
+}
+
+int create_rt_thread(pthread_t *pth, void*(*func)(void *), void *arg,
+ int policy, int prio)
+{
+ int ret;
+ struct sched_param schedp;
+ pthread_attr_t attr;
+
+ pthread_attr_init(&attr);
+ memset(&schedp, 0, sizeof(schedp));
+
+ ret = pthread_attr_setinheritsched(&attr, PTHREAD_EXPLICIT_SCHED);
+ if (ret) {
+ error("pthread_attr_setinheritsched\n", ret);
+ return -1;
+ }
+
+ ret = pthread_attr_setschedpolicy(&attr, policy);
+ if (ret) {
+ error("pthread_attr_setschedpolicy\n", ret);
+ return -1;
+ }
+
+ schedp.sched_priority = prio;
+ ret = pthread_attr_setschedparam(&attr, &schedp);
+ if (ret) {
+ error("pthread_attr_setschedparam\n", ret);
+ return -1;
+ }
+
+ ret = pthread_create(pth, &attr, func, arg);
+ if (ret) {
+ error("pthread_create\n", ret);
+ return -1;
+ }
+ return 0;
+}
+
+
+void *waiterfn(void *arg)
+{
+ struct thread_arg *args = (struct thread_arg *)arg;
+ futex_t old_val;
+
+ info("Waiter %ld: running\n", args->id);
+ /* Each thread sleeps for a different amount of time
+ * This is to avoid races, because we don't lock the
+ * external mutex here */
+ usleep(1000 * (long)args->id);
+
+ old_val = f1;
+ atomic_inc(&waiters_blocked);
+ info("Calling futex_wait_requeue_pi: %p (%u) -> %p\n",
+ &f1, f1, &f2);
+ args->ret = futex_wait_requeue_pi(&f1, old_val, &f2, args->timeout,
+ FUTEX_PRIVATE_FLAG);
+
+ info("waiter %ld woke with %d %s\n", args->id, args->ret,
+ args->ret < 0 ? strerror(errno) : "");
+ atomic_inc(&waiters_woken);
+ if (args->ret < 0) {
+ if (args->timeout && errno == ETIMEDOUT)
+ args->ret = 0;
+ else {
+ args->ret = RET_ERROR;
+ error("futex_wait_requeue_pi\n", errno);
+ }
+ futex_lock_pi(&f2, NULL, 0, FUTEX_PRIVATE_FLAG);
+ }
+ futex_unlock_pi(&f2, FUTEX_PRIVATE_FLAG);
+
+ info("Waiter %ld: exiting with %d\n", args->id, args->ret);
+ pthread_exit((void *)&args->ret);
+}
+
+void *broadcast_wakerfn(void *arg)
+{
+ struct thread_arg *args = (struct thread_arg *)arg;
+ int nr_requeue = INT_MAX;
+ int task_count = 0;
+ futex_t old_val;
+ int nr_wake = 1;
+ int i = 0;
+
+ info("Waker: waiting for waiters to block\n");
+ while (waiters_blocked.val < THREAD_MAX)
+ usleep(1000);
+ usleep(1000);
+
+ info("Waker: Calling broadcast\n");
+ if (args->lock) {
+ info("Calling FUTEX_LOCK_PI on mutex=%x @ %p\n", f2, &f2);
+ futex_lock_pi(&f2, NULL, 0, FUTEX_PRIVATE_FLAG);
+ }
+ continue_requeue:
+ old_val = f1;
+ args->ret = futex_cmp_requeue_pi(&f1, old_val, &f2, nr_wake, nr_requeue,
+ FUTEX_PRIVATE_FLAG);
+ if (args->ret < 0) {
+ args->ret = RET_ERROR;
+ error("FUTEX_CMP_REQUEUE_PI failed\n", errno);
+ } else if (++i < MAX_WAKE_ITERS) {
+ task_count += args->ret;
+ if (task_count < THREAD_MAX - waiters_woken.val)
+ goto continue_requeue;
+ } else {
+ error("max broadcast iterations (%d) reached with %d/%d tasks woken or requeued\n",
+ 0, MAX_WAKE_ITERS, task_count, THREAD_MAX);
+ args->ret = RET_ERROR;
+ }
+
+ futex_wake(&wake_complete, 1, FUTEX_PRIVATE_FLAG);
+
+ if (args->lock)
+ futex_unlock_pi(&f2, FUTEX_PRIVATE_FLAG);
+
+ if (args->ret > 0)
+ args->ret = task_count;
+
+ info("Waker: exiting with %d\n", args->ret);
+ pthread_exit((void *)&args->ret);
+}
+
+void *signal_wakerfn(void *arg)
+{
+ struct thread_arg *args = (struct thread_arg *)arg;
+ unsigned int old_val;
+ int nr_requeue = 0;
+ int task_count = 0;
+ int nr_wake = 1;
+ int i = 0;
+
+ info("Waker: waiting for waiters to block\n");
+ while (waiters_blocked.val < THREAD_MAX)
+ usleep(1000);
+ usleep(1000);
+
+ while (task_count < THREAD_MAX && waiters_woken.val < THREAD_MAX) {
+ info("task_count: %d, waiters_woken: %d\n",
+ task_count, waiters_woken.val);
+ if (args->lock) {
+ info("Calling FUTEX_LOCK_PI on mutex=%x @ %p\n",
+ f2, &f2);
+ futex_lock_pi(&f2, NULL, 0, FUTEX_PRIVATE_FLAG);
+ }
+ info("Waker: Calling signal\n");
+ /* cond_signal */
+ old_val = f1;
+ args->ret = futex_cmp_requeue_pi(&f1, old_val, &f2,
+ nr_wake, nr_requeue,
+ FUTEX_PRIVATE_FLAG);
+ if (args->ret < 0)
+ args->ret = -errno;
+ info("futex: %x\n", f2);
+ if (args->lock) {
+ info("Calling FUTEX_UNLOCK_PI on mutex=%x @ %p\n",
+ f2, &f2);
+ futex_unlock_pi(&f2, FUTEX_PRIVATE_FLAG);
+ }
+ info("futex: %x\n", f2);
+ if (args->ret < 0) {
+ error("FUTEX_CMP_REQUEUE_PI failed\n", errno);
+ args->ret = RET_ERROR;
+ break;
+ }
+
+ task_count += args->ret;
+ usleep(SIGNAL_PERIOD_US);
+ i++;
+ /* we have to loop at least THREAD_MAX times */
+ if (i > MAX_WAKE_ITERS + THREAD_MAX) {
+ error("max signaling iterations (%d) reached, giving up on pending waiters.\n",
+ 0, MAX_WAKE_ITERS + THREAD_MAX);
+ args->ret = RET_ERROR;
+ break;
+ }
+ }
+
+ futex_wake(&wake_complete, 1, FUTEX_PRIVATE_FLAG);
+
+ if (args->ret >= 0)
+ args->ret = task_count;
+
+ info("Waker: exiting with %d\n", args->ret);
+ info("Waker: waiters_woken: %d\n", waiters_woken.val);
+ pthread_exit((void *)&args->ret);
+}
+
+void *third_party_blocker(void *arg)
+{
+ struct thread_arg *args = (struct thread_arg *)arg;
+ int ret2 = 0;
+
+ args->ret = futex_lock_pi(&f2, NULL, 0, FUTEX_PRIVATE_FLAG);
+ if (args->ret)
+ goto out;
+ args->ret = futex_wait(&wake_complete, wake_complete, NULL,
+ FUTEX_PRIVATE_FLAG);
+ ret2 = futex_unlock_pi(&f2, FUTEX_PRIVATE_FLAG);
+
+ out:
+ if (args->ret || ret2) {
+ error("third_party_blocker() futex error", 0);
+ args->ret = RET_ERROR;
+ }
+
+ pthread_exit((void *)&args->ret);
+}
+
+int unit_test(int broadcast, long lock, int third_party_owner, long timeout_ns)
+{
+ void *(*wakerfn)(void *) = signal_wakerfn;
+ struct thread_arg blocker_arg = THREAD_ARG_INITIALIZER;
+ struct thread_arg waker_arg = THREAD_ARG_INITIALIZER;
+ pthread_t waiter[THREAD_MAX], waker, blocker;
+ struct timespec ts, *tsp = NULL;
+ struct thread_arg args[THREAD_MAX];
+ int *waiter_ret;
+ int i, ret = RET_PASS;
+
+ if (timeout_ns) {
+ time_t secs;
+
+ info("timeout_ns = %ld\n", timeout_ns);
+ ret = clock_gettime(CLOCK_MONOTONIC, &ts);
+ secs = (ts.tv_nsec + timeout_ns) / 1000000000;
+ ts.tv_nsec = ((int64_t)ts.tv_nsec + timeout_ns) % 1000000000;
+ ts.tv_sec += secs;
+ info("ts.tv_sec = %ld\n", ts.tv_sec);
+ info("ts.tv_nsec = %ld\n", ts.tv_nsec);
+ tsp = &ts;
+ }
+
+ if (broadcast)
+ wakerfn = broadcast_wakerfn;
+
+ if (third_party_owner) {
+ if (create_rt_thread(&blocker, third_party_blocker,
+ (void *)&blocker_arg, SCHED_FIFO, 1)) {
+ error("Creating third party blocker thread failed\n",
+ errno);
+ ret = RET_ERROR;
+ goto out;
+ }
+ }
+
+ atomic_set(&waiters_woken, 0);
+ for (i = 0; i < THREAD_MAX; i++) {
+ args[i].id = i;
+ args[i].timeout = tsp;
+ info("Starting thread %d\n", i);
+ if (create_rt_thread(&waiter[i], waiterfn, (void *)&args[i],
+ SCHED_FIFO, 1)) {
+ error("Creating waiting thread failed\n", errno);
+ ret = RET_ERROR;
+ goto out;
+ }
+ }
+ waker_arg.lock = lock;
+ if (create_rt_thread(&waker, wakerfn, (void *)&waker_arg,
+ SCHED_FIFO, 1)) {
+ error("Creating waker thread failed\n", errno);
+ ret = RET_ERROR;
+ goto out;
+ }
+
+ /* Wait for threads to finish */
+ /* Store the first error or failure encountered in waiter_ret */
+ waiter_ret = &args[0].ret;
+ for (i = 0; i < THREAD_MAX; i++)
+ pthread_join(waiter[i],
+ *waiter_ret ? NULL : (void **)&waiter_ret);
+
+ if (third_party_owner)
+ pthread_join(blocker, NULL);
+ pthread_join(waker, NULL);
+
+out:
+ if (!ret) {
+ if (*waiter_ret)
+ ret = *waiter_ret;
+ else if (waker_arg.ret < 0)
+ ret = waker_arg.ret;
+ else if (blocker_arg.ret)
+ ret = blocker_arg.ret;
+ }
+
+ return ret;
+}
+
+int main(int argc, char *argv[])
+{
+ int c, ret;
+
+ while ((c = getopt(argc, argv, "bchlot:v:")) != -1) {
+ switch (c) {
+ case 'b':
+ broadcast = 1;
+ break;
+ case 'c':
+ log_color(1);
+ break;
+ case 'h':
+ usage(basename(argv[0]));
+ exit(0);
+ case 'l':
+ locked = 1;
+ break;
+ case 'o':
+ owner = 1;
+ locked = 0;
+ break;
+ case 't':
+ timeout_ns = atoi(optarg);
+ break;
+ case 'v':
+ log_verbosity(atoi(optarg));
+ break;
+ default:
+ usage(basename(argv[0]));
+ exit(1);
+ }
+ }
+
+ printf("%s: Test requeue functionality\n", basename(argv[0]));
+ printf("\tArguments: broadcast=%d locked=%d owner=%d timeout=%ldns\n",
+ broadcast, locked, owner, timeout_ns);
+
+ /*
+ * FIXME: unit_test is obsolete now that we parse options and the
+ * various style of runs are done by run.sh - simplify the code and move
+ * unit_test into main()
+ */
+ ret = unit_test(broadcast, locked, owner, timeout_ns);
+
+ print_result(ret);
+ return ret;
+}
diff --git a/tools/testing/selftests/futex/functional/futex_requeue_pi_mismatched_ops.c b/tools/testing/selftests/futex/functional/futex_requeue_pi_mismatched_ops.c
new file mode 100644
index 0000000..d5e4f2c
--- /dev/null
+++ b/tools/testing/selftests/futex/functional/futex_requeue_pi_mismatched_ops.c
@@ -0,0 +1,135 @@
+/******************************************************************************
+ *
+ * Copyright © International Business Machines Corp., 2009
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * DESCRIPTION
+ * 1. Block a thread using FUTEX_WAIT
+ * 2. Attempt to use FUTEX_CMP_REQUEUE_PI on the futex from 1.
+ * 3. The kernel must detect the mismatch and return -EINVAL.
+ *
+ * AUTHOR
+ * Darren Hart <dvhart@linux.intel.com>
+ *
+ * HISTORY
+ * 2009-Nov-9: Initial version by Darren Hart <dvhart@linux.intel.com>
+ *
+ *****************************************************************************/
+
+#include <errno.h>
+#include <getopt.h>
+#include <pthread.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <time.h>
+#include "futextest.h"
+#include "logging.h"
+
+futex_t f1 = FUTEX_INITIALIZER;
+futex_t f2 = FUTEX_INITIALIZER;
+int child_ret = 0;
+
+void usage(char *prog)
+{
+ printf("Usage: %s\n", prog);
+ printf(" -c Use color\n");
+ printf(" -h Display this help message\n");
+ printf(" -v L Verbosity level: %d=QUIET %d=CRITICAL %d=INFO\n",
+ VQUIET, VCRITICAL, VINFO);
+}
+
+void *blocking_child(void *arg)
+{
+ child_ret = futex_wait(&f1, f1, NULL, FUTEX_PRIVATE_FLAG);
+ if (child_ret < 0) {
+ child_ret = -errno;
+ error("futex_wait\n", errno);
+ }
+ return (void *)&child_ret;
+}
+
+int main(int argc, char *argv[])
+{
+ int ret = RET_PASS;
+ pthread_t child;
+ int c;
+
+ while ((c = getopt(argc, argv, "chv:")) != -1) {
+ switch (c) {
+ case 'c':
+ log_color(1);
+ break;
+ case 'h':
+ usage(basename(argv[0]));
+ exit(0);
+ case 'v':
+ log_verbosity(atoi(optarg));
+ break;
+ default:
+ usage(basename(argv[0]));
+ exit(1);
+ }
+ }
+
+ printf("%s: Detect mismatched requeue_pi operations\n",
+ basename(argv[0]));
+
+ if (pthread_create(&child, NULL, blocking_child, NULL)) {
+ error("pthread_create\n", errno);
+ ret = RET_ERROR;
+ goto out;
+ }
+ /* Allow the child to block in the kernel. */
+ sleep(1);
+
+ /*
+ * The kernel should detect the waiter did not setup the
+ * q->requeue_pi_key and return -EINVAL. If it does not,
+ * it likely gave the lock to the child, which is now hung
+ * in the kernel.
+ */
+ ret = futex_cmp_requeue_pi(&f1, f1, &f2, 1, 0, FUTEX_PRIVATE_FLAG);
+ if (ret < 0) {
+ if (errno == EINVAL) {
+ /*
+ * The kernel correctly detected the mismatched
+ * requeue_pi target and aborted. Wake the child with
+ * FUTEX_WAKE.
+ */
+ ret = futex_wake(&f1, 1, FUTEX_PRIVATE_FLAG);
+ if (ret == 1) {
+ ret = RET_PASS;
+ } else if (ret < 0) {
+ error("futex_wake\n", errno);
+ ret = RET_ERROR;
+ } else {
+ error("futex_wake did not wake the child\n", 0);
+ ret = RET_ERROR;
+ }
+ } else {
+ error("futex_cmp_requeue_pi\n", errno);
+ ret = RET_ERROR;
+ }
+ } else if (ret > 0) {
+ fail("futex_cmp_requeue_pi failed to detect the mismatch\n");
+ ret = RET_FAIL;
+ } else {
+ error("futex_cmp_requeue_pi found no waiters\n", 0);
+ ret = RET_ERROR;
+ }
+
+ pthread_join(child, NULL);
+
+ if (!ret)
+ ret = child_ret;
+
+ out:
+ /* If the kernel crashes, we shouldn't return at all. */
+ print_result(ret);
+ return ret;
+}
diff --git a/tools/testing/selftests/futex/functional/futex_requeue_pi_signal_restart.c b/tools/testing/selftests/futex/functional/futex_requeue_pi_signal_restart.c
new file mode 100644
index 0000000..7f0c756
--- /dev/null
+++ b/tools/testing/selftests/futex/functional/futex_requeue_pi_signal_restart.c
@@ -0,0 +1,223 @@
+/******************************************************************************
+ *
+ * Copyright © International Business Machines Corp., 2006-2008
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * DESCRIPTION
+ * This test exercises the futex_wait_requeue_pi() signal handling both
+ * before and after the requeue. The first should be restarted by the
+ * kernel. The latter should return EWOULDBLOCK to the waiter.
+ *
+ * AUTHORS
+ * Darren Hart <dvhart@linux.intel.com>
+ *
+ * HISTORY
+ * 2008-May-5: Initial version by Darren Hart <dvhart@linux.intel.com>
+ *
+ *****************************************************************************/
+
+#include <errno.h>
+#include <getopt.h>
+#include <limits.h>
+#include <pthread.h>
+#include <signal.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include "atomic.h"
+#include "futextest.h"
+#include "logging.h"
+
+#define DELAY_US 100
+
+futex_t f1 = FUTEX_INITIALIZER;
+futex_t f2 = FUTEX_INITIALIZER;
+atomic_t requeued = ATOMIC_INITIALIZER;
+
+int waiter_ret = 0;
+
+void usage(char *prog)
+{
+ printf("Usage: %s\n", prog);
+ printf(" -c Use color\n");
+ printf(" -h Display this help message\n");
+ printf(" -v L Verbosity level: %d=QUIET %d=CRITICAL %d=INFO\n",
+ VQUIET, VCRITICAL, VINFO);
+}
+
+int create_rt_thread(pthread_t *pth, void*(*func)(void *), void *arg,
+ int policy, int prio)
+{
+ struct sched_param schedp;
+ pthread_attr_t attr;
+ int ret;
+
+ pthread_attr_init(&attr);
+ memset(&schedp, 0, sizeof(schedp));
+
+ ret = pthread_attr_setinheritsched(&attr, PTHREAD_EXPLICIT_SCHED);
+ if (ret) {
+ error("pthread_attr_setinheritsched\n", ret);
+ return -1;
+ }
+
+ ret = pthread_attr_setschedpolicy(&attr, policy);
+ if (ret) {
+ error("pthread_attr_setschedpolicy\n", ret);
+ return -1;
+ }
+
+ schedp.sched_priority = prio;
+ ret = pthread_attr_setschedparam(&attr, &schedp);
+ if (ret) {
+ error("pthread_attr_setschedparam\n", ret);
+ return -1;
+ }
+
+ ret = pthread_create(pth, &attr, func, arg);
+ if (ret) {
+ error("pthread_create\n", ret);
+ return -1;
+ }
+ return 0;
+}
+
+void handle_signal(int signo)
+{
+ info("signal received %s requeue\n",
+ requeued.val ? "after" : "prior to");
+}
+
+void *waiterfn(void *arg)
+{
+ unsigned int old_val;
+ int res;
+
+ waiter_ret = RET_PASS;
+
+ info("Waiter running\n");
+ info("Calling FUTEX_LOCK_PI on f2=%x @ %p\n", f2, &f2);
+ old_val = f1;
+ res = futex_wait_requeue_pi(&f1, old_val, &(f2), NULL,
+ FUTEX_PRIVATE_FLAG);
+ if (!requeued.val || errno != EWOULDBLOCK) {
+ fail("unexpected return from futex_wait_requeue_pi: %d (%s)\n",
+ res, strerror(errno));
+ info("w2:futex: %x\n", f2);
+ if (!res)
+ futex_unlock_pi(&f2, FUTEX_PRIVATE_FLAG);
+ waiter_ret = RET_FAIL;
+ }
+
+ info("Waiter exiting with %d\n", waiter_ret);
+ pthread_exit(NULL);
+}
+
+
+int main(int argc, char *argv[])
+{
+ unsigned int old_val;
+ struct sigaction sa;
+ pthread_t waiter;
+ int c, res, ret = RET_PASS;
+
+ while ((c = getopt(argc, argv, "chv:")) != -1) {
+ switch (c) {
+ case 'c':
+ log_color(1);
+ break;
+ case 'h':
+ usage(basename(argv[0]));
+ exit(0);
+ case 'v':
+ log_verbosity(atoi(optarg));
+ break;
+ default:
+ usage(basename(argv[0]));
+ exit(1);
+ }
+ }
+
+ printf("%s: Test signal handling during requeue_pi\n",
+ basename(argv[0]));
+ printf("\tArguments: <none>\n");
+
+ sa.sa_handler = handle_signal;
+ sigemptyset(&sa.sa_mask);
+ sa.sa_flags = 0;
+ if (sigaction(SIGUSR1, &sa, NULL)) {
+ error("sigaction\n", errno);
+ exit(1);
+ }
+
+ info("m1:f2: %x\n", f2);
+ info("Creating waiter\n");
+ res = create_rt_thread(&waiter, waiterfn, NULL, SCHED_FIFO, 1);
+ if (res) {
+ error("Creating waiting thread failed", res);
+ ret = RET_ERROR;
+ goto out;
+ }
+
+ info("Calling FUTEX_LOCK_PI on f2=%x @ %p\n", f2, &f2);
+ info("m2:f2: %x\n", f2);
+ futex_lock_pi(&f2, 0, 0, FUTEX_PRIVATE_FLAG);
+ info("m3:f2: %x\n", f2);
+
+ while (1) {
+ /*
+ * signal the waiter before requeue, waiter should automatically
+ * restart futex_wait_requeue_pi() in the kernel. Wait for the
+ * waiter to block on f1 again.
+ */
+ info("Issuing SIGUSR1 to waiter\n");
+ pthread_kill(waiter, SIGUSR1);
+ usleep(DELAY_US);
+
+ info("Requeueing waiter via FUTEX_CMP_REQUEUE_PI\n");
+ old_val = f1;
+ res = futex_cmp_requeue_pi(&f1, old_val, &(f2), 1, 0,
+ FUTEX_PRIVATE_FLAG);
+ /*
+ * If res is non-zero, we either requeued the waiter or hit an
+ * error, break out and handle it. If it is zero, then the
+ * signal may have hit before the the waiter was blocked on f1.
+ * Try again.
+ */
+ if (res > 0) {
+ atomic_set(&requeued, 1);
+ break;
+ } else if (res > 0) {
+ error("FUTEX_CMP_REQUEUE_PI failed\n", errno);
+ ret = RET_ERROR;
+ break;
+ }
+ }
+ info("m4:f2: %x\n", f2);
+
+ /*
+ * Signal the waiter after requeue, waiter should return from
+ * futex_wait_requeue_pi() with EWOULDBLOCK. Join the thread here so the
+ * futex_unlock_pi() can't happen before the signal wakeup is detected
+ * in the kernel.
+ */
+ info("Issuing SIGUSR1 to waiter\n");
+ pthread_kill(waiter, SIGUSR1);
+ info("Waiting for waiter to return\n");
+ pthread_join(waiter, NULL);
+
+ info("Calling FUTEX_UNLOCK_PI on mutex=%x @ %p\n", f2, &f2);
+ futex_unlock_pi(&f2, FUTEX_PRIVATE_FLAG);
+ info("m5:f2: %x\n", f2);
+
+ out:
+ if (ret == RET_PASS && waiter_ret)
+ ret = waiter_ret;
+
+ print_result(ret);
+ return ret;
+}
diff --git a/tools/testing/selftests/futex/functional/futex_wait_private_mapped_file.c b/tools/testing/selftests/futex/functional/futex_wait_private_mapped_file.c
new file mode 100644
index 0000000..5f687f2
--- /dev/null
+++ b/tools/testing/selftests/futex/functional/futex_wait_private_mapped_file.c
@@ -0,0 +1,125 @@
+/******************************************************************************
+ *
+ * Copyright FUJITSU LIMITED 2010
+ * Copyright KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * DESCRIPTION
+ * Internally, Futex has two handling mode, anon and file. The private file
+ * mapping is special. At first it behave as file, but after write anything
+ * it behave as anon. This test is intent to test such case.
+ *
+ * AUTHOR
+ * KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
+ *
+ * HISTORY
+ * 2010-Jan-6: Initial version by KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
+ *
+ *****************************************************************************/
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <syscall.h>
+#include <unistd.h>
+#include <errno.h>
+#include <linux/futex.h>
+#include <pthread.h>
+#include <libgen.h>
+#include <signal.h>
+
+#include "logging.h"
+#include "futextest.h"
+
+#define PAGE_SZ 4096
+
+char pad[PAGE_SZ] = {1};
+futex_t val = 1;
+char pad2[PAGE_SZ] = {1};
+
+#define WAKE_WAIT_US 3000000
+struct timespec wait_timeout = { .tv_sec = 5, .tv_nsec = 0};
+
+void usage(char *prog)
+{
+ printf("Usage: %s\n", prog);
+ printf(" -c Use color\n");
+ printf(" -h Display this help message\n");
+ printf(" -v L Verbosity level: %d=QUIET %d=CRITICAL %d=INFO\n",
+ VQUIET, VCRITICAL, VINFO);
+}
+
+void *thr_futex_wait(void *arg)
+{
+ int ret;
+
+ info("futex wait\n");
+ ret = futex_wait(&val, 1, &wait_timeout, 0);
+ if (ret && errno != EWOULDBLOCK && errno != ETIMEDOUT) {
+ error("futex error.\n", errno);
+ print_result(RET_ERROR);
+ exit(RET_ERROR);
+ }
+
+ if (ret && errno == ETIMEDOUT)
+ fail("waiter timedout\n");
+
+ info("futex_wait: ret = %d, errno = %d\n", ret, errno);
+
+ return NULL;
+}
+
+int main(int argc, char **argv)
+{
+ pthread_t thr;
+ int ret = RET_PASS;
+ int res;
+ int c;
+
+ while ((c = getopt(argc, argv, "chv:")) != -1) {
+ switch (c) {
+ case 'c':
+ log_color(1);
+ break;
+ case 'h':
+ usage(basename(argv[0]));
+ exit(0);
+ case 'v':
+ log_verbosity(atoi(optarg));
+ break;
+ default:
+ usage(basename(argv[0]));
+ exit(1);
+ }
+ }
+
+ printf("%s: Test the futex value of private file mappings in FUTEX_WAIT\n",
+ basename(argv[0]));
+
+ ret = pthread_create(&thr, NULL, thr_futex_wait, NULL);
+ if (ret < 0) {
+ fprintf(stderr, "pthread_create error\n");
+ ret = RET_ERROR;
+ goto out;
+ }
+
+ info("wait a while\n");
+ usleep(WAKE_WAIT_US);
+ val = 2;
+ res = futex_wake(&val, 1, 0);
+ info("futex_wake %d\n", res);
+ if (res != 1) {
+ fail("FUTEX_WAKE didn't find the waiting thread.\n");
+ ret = RET_FAIL;
+ }
+
+ info("join\n");
+ pthread_join(thr, NULL);
+
+ out:
+ print_result(ret);
+ return ret;
+}
diff --git a/tools/testing/selftests/futex/functional/futex_wait_timeout.c b/tools/testing/selftests/futex/functional/futex_wait_timeout.c
new file mode 100644
index 0000000..ab428ca
--- /dev/null
+++ b/tools/testing/selftests/futex/functional/futex_wait_timeout.c
@@ -0,0 +1,86 @@
+/******************************************************************************
+ *
+ * Copyright © International Business Machines Corp., 2009
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * DESCRIPTION
+ * Block on a futex and wait for timeout.
+ *
+ * AUTHOR
+ * Darren Hart <dvhart@linux.intel.com>
+ *
+ * HISTORY
+ * 2009-Nov-6: Initial version by Darren Hart <dvhart@linux.intel.com>
+ *
+ *****************************************************************************/
+
+#include <errno.h>
+#include <getopt.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <time.h>
+#include "futextest.h"
+#include "logging.h"
+
+static long timeout_ns = 100000; /* 100us default timeout */
+
+void usage(char *prog)
+{
+ printf("Usage: %s\n", prog);
+ printf(" -c Use color\n");
+ printf(" -h Display this help message\n");
+ printf(" -t N Timeout in nanoseconds (default: 100,000)\n");
+ printf(" -v L Verbosity level: %d=QUIET %d=CRITICAL %d=INFO\n",
+ VQUIET, VCRITICAL, VINFO);
+}
+
+int main(int argc, char *argv[])
+{
+ futex_t f1 = FUTEX_INITIALIZER;
+ struct timespec to;
+ int res, ret = RET_PASS;
+ int c;
+
+ while ((c = getopt(argc, argv, "cht:v:")) != -1) {
+ switch (c) {
+ case 'c':
+ log_color(1);
+ break;
+ case 'h':
+ usage(basename(argv[0]));
+ exit(0);
+ case 't':
+ timeout_ns = atoi(optarg);
+ break;
+ case 'v':
+ log_verbosity(atoi(optarg));
+ break;
+ default:
+ usage(basename(argv[0]));
+ exit(1);
+ }
+ }
+
+ printf("%s: Block on a futex and wait for timeout\n",
+ basename(argv[0]));
+ printf("\tArguments: timeout=%ldns\n", timeout_ns);
+
+ /* initialize timeout */
+ to.tv_sec = 0;
+ to.tv_nsec = timeout_ns;
+
+ info("Calling futex_wait on f1: %u @ %p\n", f1, &f1);
+ res = futex_wait(&f1, f1, &to, FUTEX_PRIVATE_FLAG);
+ if (!res || errno != ETIMEDOUT) {
+ fail("futex_wait returned %d\n", ret < 0 ? errno : ret);
+ ret = RET_FAIL;
+ }
+
+ print_result(ret);
+ return ret;
+}
diff --git a/tools/testing/selftests/futex/functional/futex_wait_uninitialized_heap.c b/tools/testing/selftests/futex/functional/futex_wait_uninitialized_heap.c
new file mode 100644
index 0000000..fe7aee9
--- /dev/null
+++ b/tools/testing/selftests/futex/functional/futex_wait_uninitialized_heap.c
@@ -0,0 +1,124 @@
+/******************************************************************************
+ *
+ * Copyright FUJITSU LIMITED 2010
+ * Copyright KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * DESCRIPTION
+ * Wait on uninitialized heap. It shold be zero and FUTEX_WAIT should
+ * return immediately. This test is intent to test zero page handling in
+ * futex.
+ *
+ * AUTHOR
+ * KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
+ *
+ * HISTORY
+ * 2010-Jan-6: Initial version by KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
+ *
+ *****************************************************************************/
+
+#include <pthread.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <sys/mman.h>
+#include <syscall.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <unistd.h>
+#include <errno.h>
+#include <linux/futex.h>
+#include <libgen.h>
+
+#include "logging.h"
+#include "futextest.h"
+
+#define WAIT_US 5000000
+
+static int child_blocked = 1;
+static int child_ret;
+void *buf;
+
+void usage(char *prog)
+{
+ printf("Usage: %s\n", prog);
+ printf(" -c Use color\n");
+ printf(" -h Display this help message\n");
+ printf(" -v L Verbosity level: %d=QUIET %d=CRITICAL %d=INFO\n",
+ VQUIET, VCRITICAL, VINFO);
+}
+
+void *wait_thread(void *arg)
+{
+ int res;
+
+ child_ret = RET_PASS;
+ res = futex_wait(buf, 1, NULL, 0);
+ child_blocked = 0;
+
+ if (res != 0 && errno != EWOULDBLOCK) {
+ error("futex failure\n", errno);
+ child_ret = RET_ERROR;
+ }
+ pthread_exit(NULL);
+}
+
+int main(int argc, char **argv)
+{
+ int c, ret = RET_PASS;
+ long page_size;
+ pthread_t thr;
+
+ while ((c = getopt(argc, argv, "chv:")) != -1) {
+ switch (c) {
+ case 'c':
+ log_color(1);
+ break;
+ case 'h':
+ usage(basename(argv[0]));
+ exit(0);
+ case 'v':
+ log_verbosity(atoi(optarg));
+ break;
+ default:
+ usage(basename(argv[0]));
+ exit(1);
+ }
+ }
+
+ page_size = sysconf(_SC_PAGESIZE);
+
+ buf = mmap(NULL, page_size, PROT_READ|PROT_WRITE,
+ MAP_PRIVATE|MAP_ANONYMOUS, 0, 0);
+ if (buf == (void *)-1) {
+ error("mmap\n", errno);
+ exit(1);
+ }
+
+ printf("%s: Test the uninitialized futex value in FUTEX_WAIT\n",
+ basename(argv[0]));
+
+
+ ret = pthread_create(&thr, NULL, wait_thread, NULL);
+ if (ret) {
+ error("pthread_create\n", errno);
+ ret = RET_ERROR;
+ goto out;
+ }
+
+ info("waiting %dus for child to return\n", WAIT_US);
+ usleep(WAIT_US);
+
+ ret = child_ret;
+ if (child_blocked) {
+ fail("child blocked in kernel\n");
+ ret = RET_FAIL;
+ }
+
+ out:
+ print_result(ret);
+ return ret;
+}
diff --git a/tools/testing/selftests/futex/functional/futex_wait_wouldblock.c b/tools/testing/selftests/futex/functional/futex_wait_wouldblock.c
new file mode 100644
index 0000000..b6b0274
--- /dev/null
+++ b/tools/testing/selftests/futex/functional/futex_wait_wouldblock.c
@@ -0,0 +1,79 @@
+/******************************************************************************
+ *
+ * Copyright © International Business Machines Corp., 2009
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * DESCRIPTION
+ * Test if FUTEX_WAIT op returns -EWOULDBLOCK if the futex value differs
+ * from the expected one.
+ *
+ * AUTHOR
+ * Gowrishankar <gowrishankar.m@in.ibm.com>
+ *
+ * HISTORY
+ * 2009-Nov-14: Initial version by Gowrishankar <gowrishankar.m@in.ibm.com>
+ *
+ *****************************************************************************/
+
+#include <errno.h>
+#include <getopt.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <time.h>
+#include "futextest.h"
+#include "logging.h"
+
+#define timeout_ns 100000
+
+void usage(char *prog)
+{
+ printf("Usage: %s\n", prog);
+ printf(" -c Use color\n");
+ printf(" -h Display this help message\n");
+ printf(" -v L Verbosity level: %d=QUIET %d=CRITICAL %d=INFO\n",
+ VQUIET, VCRITICAL, VINFO);
+}
+
+int main(int argc, char *argv[])
+{
+ struct timespec to = {.tv_sec = 0, .tv_nsec = timeout_ns};
+ futex_t f1 = FUTEX_INITIALIZER;
+ int res, ret = RET_PASS;
+ int c;
+
+ while ((c = getopt(argc, argv, "cht:v:")) != -1) {
+ switch (c) {
+ case 'c':
+ log_color(1);
+ break;
+ case 'h':
+ usage(basename(argv[0]));
+ exit(0);
+ case 'v':
+ log_verbosity(atoi(optarg));
+ break;
+ default:
+ usage(basename(argv[0]));
+ exit(1);
+ }
+ }
+
+ printf("%s: Test the unexpected futex value in FUTEX_WAIT\n",
+ basename(argv[0]));
+
+ info("Calling futex_wait on f1: %u @ %p with val=%u\n", f1, &f1, f1+1);
+ res = futex_wait(&f1, f1+1, &to, FUTEX_PRIVATE_FLAG);
+ if (!res || errno != EWOULDBLOCK) {
+ fail("futex_wait returned: %d %s\n",
+ res ? errno : res, res ? strerror(errno) : "");
+ ret = RET_FAIL;
+ }
+
+ print_result(ret);
+ return ret;
+}
diff --git a/tools/testing/selftests/futex/functional/run.sh b/tools/testing/selftests/futex/functional/run.sh
new file mode 100755
index 0000000..46827a8
--- /dev/null
+++ b/tools/testing/selftests/futex/functional/run.sh
@@ -0,0 +1,80 @@
+#!/bin/sh
+
+###############################################################################
+#
+# Copyright © International Business Machines Corp., 2009
+#
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 2 of the License, or
+# (at your option) any later version.
+#
+# DESCRIPTION
+# Run tests in the current directory.
+#
+# AUTHOR
+# Darren Hart <dvhart@linux.intel.com>
+#
+# HISTORY
+# 2009-Nov-9: Initial version by Darren Hart <dvhart@linux.intel.com>
+# 2010-Jan-6: Add futex_wait_uninitialized_heap and futex_wait_private_mapped_file
+# by KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
+#
+###############################################################################
+
+# Test for a color capable console
+if [ -z "$USE_COLOR" ]; then
+ tput setf 7
+ if [ $? -eq 0 ]; then
+ USE_COLOR=1
+ tput sgr0
+ fi
+fi
+if [ "$USE_COLOR" -eq 1 ]; then
+ COLOR="-c"
+fi
+
+
+echo
+# requeue pi testing
+# without timeouts
+./futex_requeue_pi $COLOR
+./futex_requeue_pi $COLOR -b
+./futex_requeue_pi $COLOR -b -l
+./futex_requeue_pi $COLOR -b -o
+./futex_requeue_pi $COLOR -l
+./futex_requeue_pi $COLOR -o
+# with timeouts
+./futex_requeue_pi $COLOR -b -l -t 5000
+./futex_requeue_pi $COLOR -l -t 5000
+./futex_requeue_pi $COLOR -b -l -t 500000
+./futex_requeue_pi $COLOR -l -t 500000
+./futex_requeue_pi $COLOR -b -t 5000
+./futex_requeue_pi $COLOR -t 5000
+./futex_requeue_pi $COLOR -b -t 500000
+./futex_requeue_pi $COLOR -t 500000
+./futex_requeue_pi $COLOR -b -o -t 5000
+./futex_requeue_pi $COLOR -l -t 5000
+./futex_requeue_pi $COLOR -b -o -t 500000
+./futex_requeue_pi $COLOR -l -t 500000
+# with long timeout
+./futex_requeue_pi $COLOR -b -l -t 2000000000
+./futex_requeue_pi $COLOR -l -t 2000000000
+
+
+echo
+./futex_requeue_pi_mismatched_ops $COLOR
+
+echo
+./futex_requeue_pi_signal_restart $COLOR
+
+echo
+./futex_wait_timeout $COLOR
+
+echo
+./futex_wait_wouldblock $COLOR
+
+echo
+./futex_wait_uninitialized_heap $COLOR
+./futex_wait_private_mapped_file $COLOR
+
diff --git a/tools/testing/selftests/futex/include/atomic.h b/tools/testing/selftests/futex/include/atomic.h
new file mode 100644
index 0000000..f861da3
--- /dev/null
+++ b/tools/testing/selftests/futex/include/atomic.h
@@ -0,0 +1,83 @@
+/******************************************************************************
+ *
+ * Copyright © International Business Machines Corp., 2009
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * DESCRIPTION
+ * GCC atomic builtin wrappers
+ * http://gcc.gnu.org/onlinedocs/gcc-4.1.0/gcc/Atomic-Builtins.html
+ *
+ * AUTHOR
+ * Darren Hart <dvhart@linux.intel.com>
+ *
+ * HISTORY
+ * 2009-Nov-17: Initial version by Darren Hart <dvhart@linux.intel.com>
+ *
+ *****************************************************************************/
+
+#ifndef _ATOMIC_H
+#define _ATOMIC_H
+
+typedef struct {
+ volatile int val;
+} atomic_t;
+
+#define ATOMIC_INITIALIZER { 0 }
+
+/**
+ * atomic_cmpxchg() - Atomic compare and exchange
+ * @uaddr: The address of the futex to be modified
+ * @oldval: The expected value of the futex
+ * @newval: The new value to try and assign the futex
+ *
+ * Return the old value of addr->val.
+ */
+static inline int
+atomic_cmpxchg(atomic_t *addr, int oldval, int newval)
+{
+ return __sync_val_compare_and_swap(&addr->val, oldval, newval);
+}
+
+/**
+ * atomic_inc() - Atomic incrememnt
+ * @addr: Address of the variable to increment
+ *
+ * Return the new value of addr->val.
+ */
+static inline int
+atomic_inc(atomic_t *addr)
+{
+ return __sync_add_and_fetch(&addr->val, 1);
+}
+
+/**
+ * atomic_dec() - Atomic decrement
+ * @addr: Address of the variable to decrement
+ *
+ * Return the new value of addr-val.
+ */
+static inline int
+atomic_dec(atomic_t *addr)
+{
+ return __sync_sub_and_fetch(&addr->val, 1);
+}
+
+/**
+ * atomic_set() - Atomic set
+ * @addr: Address of the variable to set
+ * @newval: New value for the atomic_t
+ *
+ * Return the new value of addr->val.
+ */
+static inline int
+atomic_set(atomic_t *addr, int newval)
+{
+ addr->val = newval;
+ return newval;
+}
+
+#endif
diff --git a/tools/testing/selftests/futex/include/futextest.h b/tools/testing/selftests/futex/include/futextest.h
new file mode 100644
index 0000000..b98c3ab
--- /dev/null
+++ b/tools/testing/selftests/futex/include/futextest.h
@@ -0,0 +1,266 @@
+/******************************************************************************
+ *
+ * Copyright © International Business Machines Corp., 2009
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * DESCRIPTION
+ * Glibc independent futex library for testing kernel functionality.
+ *
+ * AUTHOR
+ * Darren Hart <dvhart@linux.intel.com>
+ *
+ * HISTORY
+ * 2009-Nov-6: Initial version by Darren Hart <dvhart@linux.intel.com>
+ *
+ *****************************************************************************/
+
+#ifndef _FUTEXTEST_H
+#define _FUTEXTEST_H
+
+#include <unistd.h>
+#include <sys/syscall.h>
+#include <sys/types.h>
+#include <linux/futex.h>
+
+typedef volatile u_int32_t futex_t;
+#define FUTEX_INITIALIZER 0
+
+/* Define the newer op codes if the system header file is not up to date. */
+#ifndef FUTEX_WAIT_BITSET
+#define FUTEX_WAIT_BITSET 9
+#endif
+#ifndef FUTEX_WAKE_BITSET
+#define FUTEX_WAKE_BITSET 10
+#endif
+#ifndef FUTEX_WAIT_REQUEUE_PI
+#define FUTEX_WAIT_REQUEUE_PI 11
+#endif
+#ifndef FUTEX_CMP_REQUEUE_PI
+#define FUTEX_CMP_REQUEUE_PI 12
+#endif
+#ifndef FUTEX_WAIT_REQUEUE_PI_PRIVATE
+#define FUTEX_WAIT_REQUEUE_PI_PRIVATE (FUTEX_WAIT_REQUEUE_PI | \
+ FUTEX_PRIVATE_FLAG)
+#endif
+#ifndef FUTEX_REQUEUE_PI_PRIVATE
+#define FUTEX_CMP_REQUEUE_PI_PRIVATE (FUTEX_CMP_REQUEUE_PI | \
+ FUTEX_PRIVATE_FLAG)
+#endif
+
+/**
+ * futex() - SYS_futex syscall wrapper
+ * @uaddr: address of first futex
+ * @op: futex op code
+ * @val: typically expected value of uaddr, but varies by op
+ * @timeout: typically an absolute struct timespec (except where noted
+ * otherwise). Overloaded by some ops
+ * @uaddr2: address of second futex for some ops\
+ * @val3: varies by op
+ * @opflags: flags to be bitwise OR'd with op, such as FUTEX_PRIVATE_FLAG
+ *
+ * futex() is used by all the following futex op wrappers. It can also be
+ * used for misuse and abuse testing. Generally, the specific op wrappers
+ * should be used instead. It is a macro instead of an static inline function as
+ * some of the types over overloaded (timeout is used for nr_requeue for
+ * example).
+ *
+ * These argument descriptions are the defaults for all
+ * like-named arguments in the following wrappers except where noted below.
+ */
+#define futex(uaddr, op, val, timeout, uaddr2, val3, opflags) \
+ syscall(SYS_futex, uaddr, op | opflags, val, timeout, uaddr2, val3)
+
+/**
+ * futex_wait() - block on uaddr with optional timeout
+ * @timeout: relative timeout
+ */
+static inline int
+futex_wait(futex_t *uaddr, futex_t val, struct timespec *timeout, int opflags)
+{
+ return futex(uaddr, FUTEX_WAIT, val, timeout, NULL, 0, opflags);
+}
+
+/**
+ * futex_wake() - wake one or more tasks blocked on uaddr
+ * @nr_wake: wake up to this many tasks
+ */
+static inline int
+futex_wake(futex_t *uaddr, int nr_wake, int opflags)
+{
+ return futex(uaddr, FUTEX_WAKE, nr_wake, NULL, NULL, 0, opflags);
+}
+
+/**
+ * futex_wait_bitset() - block on uaddr with bitset
+ * @bitset: bitset to be used with futex_wake_bitset
+ */
+static inline int
+futex_wait_bitset(futex_t *uaddr, futex_t val, struct timespec *timeout,
+ u_int32_t bitset, int opflags)
+{
+ return futex(uaddr, FUTEX_WAIT_BITSET, val, timeout, NULL, bitset,
+ opflags);
+}
+
+/**
+ * futex_wake_bitset() - wake one or more tasks blocked on uaddr with bitset
+ * @bitset: bitset to compare with that used in futex_wait_bitset
+ */
+static inline int
+futex_wake_bitset(futex_t *uaddr, int nr_wake, u_int32_t bitset, int opflags)
+{
+ return futex(uaddr, FUTEX_WAKE_BITSET, nr_wake, NULL, NULL, bitset,
+ opflags);
+}
+
+/**
+ * futex_lock_pi() - block on uaddr as a PI mutex
+ * @detect: whether (1) or not (0) to perform deadlock detection
+ */
+static inline int
+futex_lock_pi(futex_t *uaddr, struct timespec *timeout, int detect,
+ int opflags)
+{
+ return futex(uaddr, FUTEX_LOCK_PI, detect, timeout, NULL, 0, opflags);
+}
+
+/**
+ * futex_unlock_pi() - release uaddr as a PI mutex, waking the top waiter
+ */
+static inline int
+futex_unlock_pi(futex_t *uaddr, int opflags)
+{
+ return futex(uaddr, FUTEX_UNLOCK_PI, 0, NULL, NULL, 0, opflags);
+}
+
+/**
+ * futex_wake_op() - FIXME: COME UP WITH A GOOD ONE LINE DESCRIPTION
+ */
+static inline int
+futex_wake_op(futex_t *uaddr, futex_t *uaddr2, int nr_wake, int nr_wake2,
+ int wake_op, int opflags)
+{
+ return futex(uaddr, FUTEX_WAKE_OP, nr_wake, nr_wake2, uaddr2, wake_op,
+ opflags);
+}
+
+/**
+ * futex_requeue() - requeue without expected value comparison, deprecated
+ * @nr_wake: wake up to this many tasks
+ * @nr_requeue: requeue up to this many tasks
+ *
+ * Due to its inherently racy implementation, futex_requeue() is deprecated in
+ * favor of futex_cmp_requeue().
+ */
+static inline int
+futex_requeue(futex_t *uaddr, futex_t *uaddr2, int nr_wake, int nr_requeue,
+ int opflags)
+{
+ return futex(uaddr, FUTEX_REQUEUE, nr_wake, nr_requeue, uaddr2, 0,
+ opflags);
+}
+
+/**
+ * futex_cmp_requeue() - requeue tasks from uaddr to uaddr2
+ * @nr_wake: wake up to this many tasks
+ * @nr_requeue: requeue up to this many tasks
+ */
+static inline int
+futex_cmp_requeue(futex_t *uaddr, futex_t val, futex_t *uaddr2, int nr_wake,
+ int nr_requeue, int opflags)
+{
+ return futex(uaddr, FUTEX_CMP_REQUEUE, nr_wake, nr_requeue, uaddr2,
+ val, opflags);
+}
+
+/**
+ * futex_wait_requeue_pi() - block on uaddr and prepare to requeue to uaddr2
+ * @uaddr: non-PI futex source
+ * @uaddr2: PI futex target
+ *
+ * This is the first half of the requeue_pi mechanism. It shall always be
+ * paired with futex_cmp_requeue_pi().
+ */
+static inline int
+futex_wait_requeue_pi(futex_t *uaddr, futex_t val, futex_t *uaddr2,
+ struct timespec *timeout, int opflags)
+{
+ return futex(uaddr, FUTEX_WAIT_REQUEUE_PI, val, timeout, uaddr2, 0,
+ opflags);
+}
+
+/**
+ * futex_cmp_requeue_pi() - requeue tasks from uaddr to uaddr2 (PI aware)
+ * @uaddr: non-PI futex source
+ * @uaddr2: PI futex target
+ * @nr_wake: wake up to this many tasks
+ * @nr_requeue: requeue up to this many tasks
+ */
+static inline int
+futex_cmp_requeue_pi(futex_t *uaddr, futex_t val, futex_t *uaddr2, int nr_wake,
+ int nr_requeue, int opflags)
+{
+ return futex(uaddr, FUTEX_CMP_REQUEUE_PI, nr_wake, nr_requeue, uaddr2,
+ val, opflags);
+}
+
+/**
+ * futex_cmpxchg() - atomic compare and exchange
+ * @uaddr: The address of the futex to be modified
+ * @oldval: The expected value of the futex
+ * @newval: The new value to try and assign the futex
+ *
+ * Implement cmpxchg using gcc atomic builtins.
+ * http://gcc.gnu.org/onlinedocs/gcc-4.1.0/gcc/Atomic-Builtins.html
+ *
+ * Return the old futex value.
+ */
+static inline u_int32_t
+futex_cmpxchg(futex_t *uaddr, u_int32_t oldval, u_int32_t newval)
+{
+ return __sync_val_compare_and_swap(uaddr, oldval, newval);
+}
+
+/**
+ * futex_dec() - atomic decrement of the futex value
+ * @uaddr: The address of the futex to be modified
+ *
+ * Return the new futex value.
+ */
+static inline u_int32_t
+futex_dec(futex_t *uaddr)
+{
+ return __sync_sub_and_fetch(uaddr, 1);
+}
+
+/**
+ * futex_inc() - atomic increment of the futex value
+ * @uaddr: the address of the futex to be modified
+ *
+ * Return the new futex value.
+ */
+static inline u_int32_t
+futex_inc(futex_t *uaddr)
+{
+ return __sync_add_and_fetch(uaddr, 1);
+}
+
+/**
+ * futex_set() - atomic decrement of the futex value
+ * @uaddr: the address of the futex to be modified
+ * @newval: New value for the atomic_t
+ *
+ * Return the new futex value.
+ */
+static inline u_int32_t
+futex_set(futex_t *uaddr, u_int32_t newval)
+{
+ *uaddr = newval;
+ return newval;
+}
+
+#endif
diff --git a/tools/testing/selftests/futex/include/logging.h b/tools/testing/selftests/futex/include/logging.h
new file mode 100644
index 0000000..f6ed5c2
--- /dev/null
+++ b/tools/testing/selftests/futex/include/logging.h
@@ -0,0 +1,150 @@
+/******************************************************************************
+ *
+ * Copyright © International Business Machines Corp., 2009
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * DESCRIPTION
+ * Glibc independent futex library for testing kernel functionality.
+ *
+ * AUTHOR
+ * Darren Hart <dvhart@linux.intel.com>
+ *
+ * HISTORY
+ * 2009-Nov-6: Initial version by Darren Hart <dvhart@linux.intel.com>
+ *
+ *****************************************************************************/
+
+#ifndef _LOGGING_H
+#define _LOGGING_H
+
+#include <string.h>
+#include <unistd.h>
+#include <linux/futex.h>
+
+/*
+ * Define PASS, ERROR, and FAIL strings with and without color escape
+ * sequences, default to no color.
+ */
+#define ESC 0x1B, '['
+#define BRIGHT '1'
+#define GREEN '3', '2'
+#define YELLOW '3', '3'
+#define RED '3', '1'
+#define ESCEND 'm'
+#define BRIGHT_GREEN ESC, BRIGHT, ';', GREEN, ESCEND
+#define BRIGHT_YELLOW ESC, BRIGHT, ';', YELLOW, ESCEND
+#define BRIGHT_RED ESC, BRIGHT, ';', RED, ESCEND
+#define RESET_COLOR ESC, '0', 'm'
+static const char PASS_COLOR[] = {BRIGHT_GREEN, ' ', 'P', 'A', 'S', 'S',
+ RESET_COLOR, 0};
+static const char ERROR_COLOR[] = {BRIGHT_YELLOW, 'E', 'R', 'R', 'O', 'R',
+ RESET_COLOR, 0};
+static const char FAIL_COLOR[] = {BRIGHT_RED, ' ', 'F', 'A', 'I', 'L',
+ RESET_COLOR, 0};
+static const char INFO_NORMAL[] = " INFO";
+static const char PASS_NORMAL[] = " PASS";
+static const char ERROR_NORMAL[] = "ERROR";
+static const char FAIL_NORMAL[] = " FAIL";
+const char *INFO = INFO_NORMAL;
+const char *PASS = PASS_NORMAL;
+const char *ERROR = ERROR_NORMAL;
+const char *FAIL = FAIL_NORMAL;
+
+/* Verbosity setting for INFO messages */
+#define VQUIET 0
+#define VCRITICAL 1
+#define VINFO 2
+#define VMAX VINFO
+int _verbose = VCRITICAL;
+
+/* Functional test return codes */
+#define RET_PASS 0
+#define RET_ERROR -1
+#define RET_FAIL -2
+
+/**
+ * log_color() - Use colored output for PASS, ERROR, and FAIL strings
+ * @use_color: use color (1) or not (0)
+ */
+void log_color(int use_color)
+{
+ if (use_color) {
+ PASS = PASS_COLOR;
+ ERROR = ERROR_COLOR;
+ FAIL = FAIL_COLOR;
+ } else {
+ PASS = PASS_NORMAL;
+ ERROR = ERROR_NORMAL;
+ FAIL = FAIL_NORMAL;
+ }
+}
+
+/**
+ * log_verbosity() - Set verbosity of test output
+ * @verbose: Enable (1) verbose output or not (0)
+ *
+ * Currently setting verbose=1 will enable INFO messages and 0 will disable
+ * them. FAIL and ERROR messages are always displayed.
+ */
+void log_verbosity(int level)
+{
+ if (level > VMAX)
+ level = VMAX;
+ else if (level < 0)
+ level = 0;
+ _verbose = level;
+}
+
+/**
+ * print_result() - Print standard PASS | ERROR | FAIL results
+ * @ret: the return value to be considered: 0 | RET_ERROR | RET_FAIL
+ *
+ * print_result() is primarily intended for functional tests.
+ */
+void print_result(int ret)
+{
+ const char *result = "Unknown return code";
+
+ switch (ret) {
+ case RET_PASS:
+ result = PASS;
+ break;
+ case RET_ERROR:
+ result = ERROR;
+ break;
+ case RET_FAIL:
+ result = FAIL;
+ break;
+ }
+ printf("Result: %s\n", result);
+}
+
+/* log level macros */
+#define info(message, vargs...) \
+do { \
+ if (_verbose >= VINFO) \
+ fprintf(stderr, "\t%s: "message, INFO, ##vargs); \
+} while (0)
+
+#define error(message, err, args...) \
+do { \
+ if (_verbose >= VCRITICAL) {\
+ if (err) \
+ fprintf(stderr, "\t%s: %s: "message, \
+ ERROR, strerror(err), ##args); \
+ else \
+ fprintf(stderr, "\t%s: "message, ERROR, ##args); \
+ } \
+} while (0)
+
+#define fail(message, args...) \
+do { \
+ if (_verbose >= VCRITICAL) \
+ fprintf(stderr, "\t%s: "message, FAIL, ##args); \
+} while (0)
+
+#endif
diff --git a/tools/testing/selftests/futex/run.sh b/tools/testing/selftests/futex/run.sh
new file mode 100755
index 0000000..4126312
--- /dev/null
+++ b/tools/testing/selftests/futex/run.sh
@@ -0,0 +1,33 @@
+#!/bin/sh
+
+###############################################################################
+#
+# Copyright © International Business Machines Corp., 2009
+#
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 2 of the License, or
+# (at your option) any later version.
+#
+# DESCRIPTION
+# Run all tests under the functional, performance, and stress directories.
+# Format and summarize the results.
+#
+# AUTHOR
+# Darren Hart <dvhart@linux.intel.com>
+#
+# HISTORY
+# 2009-Nov-9: Initial version by Darren Hart <dvhart@linux.intel.com>
+#
+###############################################################################
+
+# Test for a color capable shell and pass the result to the subdir scripts
+USE_COLOR=0
+tput setf 7
+if [ $? -eq 0 ]; then
+ USE_COLOR=1
+ tput sgr0
+fi
+export USE_COLOR
+
+(cd functional; ./run.sh)
--
2.1.4
^ permalink raw reply related
* [PATCH 2/5] selftest/futex: Update Makefile to use lib.mk
From: Darren Hart @ 2015-05-08 22:09 UTC (permalink / raw)
To: linux-api-u79uwXL29TY76Z2rM5mHXA, Linux Kernel Mailing List
Cc: Darren Hart, Shuah Khan, Ingo Molnar, Peter Zijlstra,
Thomas Gleixner, Davidlohr Bueso, KOSAKI Motohiro
In-Reply-To: <cover.1431121818.git.dvhart-VuQAYsv1563Yd54FQh9/CA@public.gmane.org>
Adapt the futextest Makefiles to use lib.mk macros for RUN_TESTS and
EMIT_TESTS. For now, we reuse the run.sh mechanism provided by
futextest. This doesn't provide the standard selftests: [PASS|FAIL]
format, but the tests provide very similar output already.
This results in the run_kselftest.sh script for futexes including a
single line: ./run.sh
Cc: Shuah Khan <shuahkh-JPH+aEBZ4P+UEJcrhfAQsw@public.gmane.org>
Cc: linux-api-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
Cc: Ingo Molnar <mingo-X9Un+BFzKDI@public.gmane.org>
Cc: Peter Zijlstra <peterz-wEGCiKHe2LqWVfeAwA7xHQ@public.gmane.org>
Cc: Thomas Gleixner <tglx-hfZtesqFncYOwBW4kG4KsQ@public.gmane.org>
Cc: Davidlohr Bueso <dave-h16yJtLeMjHk1uMJSBkQmQ@public.gmane.org>
Cc: KOSAKI Motohiro <kosaki.motohiro-+CUm20s59erQFUHtdCDX3A@public.gmane.org>
Signed-off-by: Darren Hart <dvhart-VuQAYsv1563Yd54FQh9/CA@public.gmane.org>
---
tools/testing/selftests/futex/Makefile | 20 +++++++++++++++++++-
tools/testing/selftests/futex/functional/Makefile | 5 +++--
2 files changed, 22 insertions(+), 3 deletions(-)
diff --git a/tools/testing/selftests/futex/Makefile b/tools/testing/selftests/futex/Makefile
index 2c26d59..6a17529 100644
--- a/tools/testing/selftests/futex/Makefile
+++ b/tools/testing/selftests/futex/Makefile
@@ -1,11 +1,29 @@
SUBDIRS := functional
+TEST_PROGS := run.sh
+
.PHONY: all clean
all:
for DIR in $(SUBDIRS); do $(MAKE) -C $$DIR $@ ; done
-run_tests: all
+include ../lib.mk
+
+override define RUN_TESTS
./run.sh
+endef
+
+override define INSTALL_RULE
+ mkdir -p $(INSTALL_PATH)
+ install -t $(INSTALL_PATH) $(TEST_PROGS) $(TEST_PROGS_EXTENDED) $(TEST_FILES)
+
+ @for SUBDIR in $(SUBDIRS); do \
+ $(MAKE) -C $$SUBDIR INSTALL_PATH=$(INSTALL_PATH)/$$SUBDIR install; \
+ done;
+endef
+
+override define EMIT_TESTS
+ echo "./run.sh"
+endef
clean:
for DIR in $(SUBDIRS); do $(MAKE) -C $$DIR $@ ; done
diff --git a/tools/testing/selftests/futex/functional/Makefile b/tools/testing/selftests/futex/functional/Makefile
index 4098340..fb96927 100644
--- a/tools/testing/selftests/futex/functional/Makefile
+++ b/tools/testing/selftests/futex/functional/Makefile
@@ -12,13 +12,14 @@ TARGETS := \
futex_wait_uninitialized_heap \
futex_wait_private_mapped_file
+TEST_PROGS := $(TARGETS) run.sh
+
.PHONY: all clean
all: $(TARGETS)
$(TARGETS): $(HEADERS)
-run_tests: all
- ./run.sh
+include ../../lib.mk
clean:
rm -f $(TARGETS)
--
2.1.4
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox