* Re: [PATCH v2 2/3] mm/memory_hotplug: Introduce MHP_NO_FIRMWARE_MEMMAP
From: Eric W. Biederman @ 2020-04-30 16:33 UTC (permalink / raw)
To: David Hildenbrand
Cc: virtio-dev, linux-hyperv, Michal Hocko, Baoquan He, linux-mm,
Wei Yang, linux-s390, linux-nvdimm, linux-kernel, virtualization,
linux-acpi, Michael S . Tsirkin, Pankaj Gupta, xen-devel,
Andrew Morton, Michal Hocko, linuxppc-dev
In-Reply-To: <1b49c3be-6e2f-57cb-96f7-f66a8f8a9380@redhat.com>
David Hildenbrand <david@redhat.com> writes:
> On 30.04.20 17:38, Eric W. Biederman wrote:
>> David Hildenbrand <david@redhat.com> writes:
>>
>>> Some devices/drivers that add memory via add_memory() and friends (e.g.,
>>> dax/kmem, but also virtio-mem in the future) don't want to create entries
>>> in /sys/firmware/memmap/ - primarily to hinder kexec from adding this
>>> memory to the boot memmap of the kexec kernel.
>>>
>>> In fact, such memory is never exposed via the firmware memmap as System
>>> RAM (e.g., e820), so exposing this memory via /sys/firmware/memmap/ is
>>> wrong:
>>> "kexec needs the raw firmware-provided memory map to setup the
>>> parameter segment of the kernel that should be booted with
>>> kexec. Also, the raw memory map is useful for debugging. For
>>> that reason, /sys/firmware/memmap is an interface that provides
>>> the raw memory map to userspace." [1]
>>>
>>> We don't have to worry about firmware_map_remove() on the removal path.
>>> If there is no entry, it will simply return with -EINVAL.
>>>
>>> [1]
>>> https://www.kernel.org/doc/Documentation/ABI/testing/sysfs-firmware-memmap
>>
>>
>> You know what this justification is rubbish, and I have previously
>> explained why it is rubbish.
>
> Actually, no, I don't think it is rubbish. See patch #3 and the cover
> letter why this is the right thing to do *for special memory*, *not
> ordinary DIMMs*.
>
> And to be quite honest, I think your response is a little harsh. I don't
> recall you replying to my virtio-mem-related comments.
>
>>
>> Nacked-by: "Eric W. Biederman" <ebiederm@xmission.com>
>>
>> This needs to be based on weather the added memory is ultimately normal
>> ram or is something special.
>
> Yes, that's what the caller are expected to decide, see patch #3.
>
> kexec should try to be as closely as possible to a real reboot - IMHO.
That is very fuzzy in terms of hotplug memory. The kexec'd kernel
should see the hotplugged memory assuming it is ordinary memory.
But kexec is not a reboot although it is quite similar. Kexec is
swapping one running kernel and it's state for another kernel without
rebooting.
>> Justifying behavior by documentation that does not consider memory
>> hotplug is bad thinking.
>
> Are you maybe confusing this patch series with the arm64 approach? This
> is not about ordinary hotplugged DIMMs.
I think I am.
My challenge is that I don't see anything in the description that says
this isn't about ordinary hotplugged DIMMs. All I saw was hotplug
memory.
If the class of memory is different then please by all means let's mark
it differently in struct resource so everyone knows it is different.
But that difference needs to be more than hotplug.
That difference needs to be the hypervisor loaned us memory and might
take it back at any time, or this memory is persistent and so it has
these different characteristics so don't use it as ordinary ram.
That information is also useful to other people looking at the system
and seeing what is going on.
Just please don't muddle the concepts, or assume that whatever subset of
hotplug memory you are dealing with is the only subset.
I didn't see that flag making the distinction about the kind of memory
it is.
Eric
^ permalink raw reply
* Re: [PATCH v2 2/3] mm/memory_hotplug: Introduce MHP_NO_FIRMWARE_MEMMAP
From: David Hildenbrand @ 2020-04-30 16:49 UTC (permalink / raw)
To: Eric W. Biederman
Cc: virtio-dev, linux-hyperv, Michal Hocko, Baoquan He, linux-mm,
Wei Yang, linux-s390, linux-nvdimm, linux-kernel, virtualization,
linux-acpi, Michael S . Tsirkin, Pankaj Gupta, xen-devel,
Andrew Morton, Michal Hocko, linuxppc-dev
In-Reply-To: <871ro52ary.fsf@x220.int.ebiederm.org>
On 30.04.20 18:33, Eric W. Biederman wrote:
> David Hildenbrand <david@redhat.com> writes:
>
>> On 30.04.20 17:38, Eric W. Biederman wrote:
>>> David Hildenbrand <david@redhat.com> writes:
>>>
>>>> Some devices/drivers that add memory via add_memory() and friends (e.g.,
>>>> dax/kmem, but also virtio-mem in the future) don't want to create entries
>>>> in /sys/firmware/memmap/ - primarily to hinder kexec from adding this
>>>> memory to the boot memmap of the kexec kernel.
>>>>
>>>> In fact, such memory is never exposed via the firmware memmap as System
>>>> RAM (e.g., e820), so exposing this memory via /sys/firmware/memmap/ is
>>>> wrong:
>>>> "kexec needs the raw firmware-provided memory map to setup the
>>>> parameter segment of the kernel that should be booted with
>>>> kexec. Also, the raw memory map is useful for debugging. For
>>>> that reason, /sys/firmware/memmap is an interface that provides
>>>> the raw memory map to userspace." [1]
>>>>
>>>> We don't have to worry about firmware_map_remove() on the removal path.
>>>> If there is no entry, it will simply return with -EINVAL.
>>>>
>>>> [1]
>>>> https://www.kernel.org/doc/Documentation/ABI/testing/sysfs-firmware-memmap
>>>
>>>
>>> You know what this justification is rubbish, and I have previously
>>> explained why it is rubbish.
>>
>> Actually, no, I don't think it is rubbish. See patch #3 and the cover
>> letter why this is the right thing to do *for special memory*, *not
>> ordinary DIMMs*.
>>
>> And to be quite honest, I think your response is a little harsh. I don't
>> recall you replying to my virtio-mem-related comments.
>>
>>>
>>> Nacked-by: "Eric W. Biederman" <ebiederm@xmission.com>
>>>
>>> This needs to be based on weather the added memory is ultimately normal
>>> ram or is something special.
>>
>> Yes, that's what the caller are expected to decide, see patch #3.
>>
>> kexec should try to be as closely as possible to a real reboot - IMHO.
>
> That is very fuzzy in terms of hotplug memory. The kexec'd kernel
> should see the hotplugged memory assuming it is ordinary memory.
>
> But kexec is not a reboot although it is quite similar. Kexec is
> swapping one running kernel and it's state for another kernel without
> rebooting.
I agree (especially regarding the arm64 DIMM hotplug discussion).
However, for the two cases
a) dax/kmem
b) virtio-mem
We really want to let the driver take back control and figure out "what
to do with the memory".
>
>>> Justifying behavior by documentation that does not consider memory
>>> hotplug is bad thinking.
>>
>> Are you maybe confusing this patch series with the arm64 approach? This
>> is not about ordinary hotplugged DIMMs.
>
> I think I am.
>
> My challenge is that I don't see anything in the description that says
> this isn't about ordinary hotplugged DIMMs. All I saw was hotplug
> memory.
I'm sorry if that was confusing, I tried to stress that kmem and
virtio-mem is special in the description.
I squeezed a lot of that information into the cover letter and into
patch #3.
>
> If the class of memory is different then please by all means let's mark
> it differently in struct resource so everyone knows it is different.
> But that difference needs to be more than hotplug.
>
> That difference needs to be the hypervisor loaned us memory and might
> take it back at any time, or this memory is persistent and so it has
> these different characteristics so don't use it as ordinary ram.
Yes, and I think kmem took an excellent approach of explicitly putting
that "System RAM" into a resource hierarchy. That "System RAM" won't
show up as a root node under /proc/iomem (see patch #3), which already
results in kexec-tools to treat it in a special way. I am thinking about
doing the same for virtio-mem.
>
> That information is also useful to other people looking at the system
> and seeing what is going on.
>
> Just please don't muddle the concepts, or assume that whatever subset of
> hotplug memory you are dealing with is the only subset.
I can certainly rephrase the subject/description/comment, stating that
this is not to be used for ordinary hotplugged DIMMs - only when the
device driver is under control to decide what to do with that memory -
especially when kexec'ing.
(previously, I called this flag MHP_DRIVER_MANAGED, but I think
MHP_NO_FIRMWARE_MEMMAP is clearer, we just need a better description)
Would that make it clearer?
Thanks!
--
Thanks,
David / dhildenb
^ permalink raw reply
* Re: [PATCH v2 2/3] mm/memory_hotplug: Introduce MHP_NO_FIRMWARE_MEMMAP
From: Eric W. Biederman @ 2020-04-30 18:06 UTC (permalink / raw)
To: David Hildenbrand
Cc: virtio-dev, linux-hyperv, Michal Hocko, Baoquan He, linux-mm,
Wei Yang, linux-s390, linux-nvdimm, linux-kernel, virtualization,
linux-acpi, Michael S . Tsirkin, Pankaj Gupta, xen-devel,
Andrew Morton, Michal Hocko, linuxppc-dev
In-Reply-To: <373a6898-4020-4af1-5b3d-f827d705dd77@redhat.com>
David Hildenbrand <david@redhat.com> writes:
> On 30.04.20 18:33, Eric W. Biederman wrote:
>> David Hildenbrand <david@redhat.com> writes:
>>
>>> On 30.04.20 17:38, Eric W. Biederman wrote:
>>>> David Hildenbrand <david@redhat.com> writes:
>>>>
>>>>> Some devices/drivers that add memory via add_memory() and friends (e.g.,
>>>>> dax/kmem, but also virtio-mem in the future) don't want to create entries
>>>>> in /sys/firmware/memmap/ - primarily to hinder kexec from adding this
>>>>> memory to the boot memmap of the kexec kernel.
>>>>>
>>>>> In fact, such memory is never exposed via the firmware memmap as System
>>>>> RAM (e.g., e820), so exposing this memory via /sys/firmware/memmap/ is
>>>>> wrong:
>>>>> "kexec needs the raw firmware-provided memory map to setup the
>>>>> parameter segment of the kernel that should be booted with
>>>>> kexec. Also, the raw memory map is useful for debugging. For
>>>>> that reason, /sys/firmware/memmap is an interface that provides
>>>>> the raw memory map to userspace." [1]
>>>>>
>>>>> We don't have to worry about firmware_map_remove() on the removal path.
>>>>> If there is no entry, it will simply return with -EINVAL.
>>>>>
>>>>> [1]
>>>>> https://www.kernel.org/doc/Documentation/ABI/testing/sysfs-firmware-memmap
>>>>
>>>>
>>>> You know what this justification is rubbish, and I have previously
>>>> explained why it is rubbish.
>>>
>>> Actually, no, I don't think it is rubbish. See patch #3 and the cover
>>> letter why this is the right thing to do *for special memory*, *not
>>> ordinary DIMMs*.
>>>
>>> And to be quite honest, I think your response is a little harsh. I don't
>>> recall you replying to my virtio-mem-related comments.
>>>
>>>>
>>>> Nacked-by: "Eric W. Biederman" <ebiederm@xmission.com>
>>>>
>>>> This needs to be based on weather the added memory is ultimately normal
>>>> ram or is something special.
>>>
>>> Yes, that's what the caller are expected to decide, see patch #3.
>>>
>>> kexec should try to be as closely as possible to a real reboot - IMHO.
>>
>> That is very fuzzy in terms of hotplug memory. The kexec'd kernel
>> should see the hotplugged memory assuming it is ordinary memory.
>>
>> But kexec is not a reboot although it is quite similar. Kexec is
>> swapping one running kernel and it's state for another kernel without
>> rebooting.
>
> I agree (especially regarding the arm64 DIMM hotplug discussion).
> However, for the two cases
>
> a) dax/kmem
> b) virtio-mem
>
> We really want to let the driver take back control and figure out "what
> to do with the memory".
From reading your v1 cover letter (the description appears missing in
v2) I see what you are talking about with respect to virtio-mem.
So I will count virt-io mem as something different.
>>>> Justifying behavior by documentation that does not consider memory
>>>> hotplug is bad thinking.
>>>
>>> Are you maybe confusing this patch series with the arm64 approach? This
>>> is not about ordinary hotplugged DIMMs.
>>
>> I think I am.
>>
>> My challenge is that I don't see anything in the description that says
>> this isn't about ordinary hotplugged DIMMs. All I saw was hotplug
>> memory.
>
> I'm sorry if that was confusing, I tried to stress that kmem and
> virtio-mem is special in the description.
>
> I squeezed a lot of that information into the cover letter and into
> patch #3.
>> If the class of memory is different then please by all means let's mark
>> it differently in struct resource so everyone knows it is different.
>> But that difference needs to be more than hotplug.
>>
>> That difference needs to be the hypervisor loaned us memory and might
>> take it back at any time, or this memory is persistent and so it has
>> these different characteristics so don't use it as ordinary ram.
>
> Yes, and I think kmem took an excellent approach of explicitly putting
> that "System RAM" into a resource hierarchy. That "System RAM" won't
> show up as a root node under /proc/iomem (see patch #3), which already
> results in kexec-tools to treat it in a special way. I am thinking about
> doing the same for virtio-mem.
Reading this and your patch cover letters again my concern is that
the justification seems to be letting the tail wag the dog.
You want kexec-tools to behave in a certain way so you are changing the
kernel.
Rather it should be change the kernel to clearly reflect reality and if
you can get away without a change to kexec-tools that is a bonus.
>> That information is also useful to other people looking at the system
>> and seeing what is going on.
>>
>> Just please don't muddle the concepts, or assume that whatever subset of
>> hotplug memory you are dealing with is the only subset.
>
> I can certainly rephrase the subject/description/comment, stating that
> this is not to be used for ordinary hotplugged DIMMs - only when the
> device driver is under control to decide what to do with that memory -
> especially when kexec'ing.
>
> (previously, I called this flag MHP_DRIVER_MANAGED, but I think
> MHP_NO_FIRMWARE_MEMMAP is clearer, we just need a better description)
>
> Would that make it clearer?
I am not certain, but Andrew Morton deliberately added that
firmware_map_add_hotplug call. Which means that there is a reason
for putting hotplugged memory in the firmware map.
So the justification needs to take that reason into account. The
justification can not be it is hotplugged therefore it should not belong
in the firmware memory map. Unless you can show that
firmware_map_add_hotplug that was actually a bug and should be removed.
But as it has been that way since 2010 that seems like a long shot.
So my question is what is right for the firmware map?
Why does the firmware map support hotplug entries?
Once we have the answers to those questions we can figure out what logic
the special kinds of memory hotplug need.
Ref: d96ae5309165 ("memory-hotplug: create /sys/firmware/memmap entry for new memory")
Eric
^ permalink raw reply
* [PATCH net] ibmvnic: Skip fatal error reset after passive init
From: Juliet Kim @ 2020-04-30 18:22 UTC (permalink / raw)
To: netdev; +Cc: julietk, tlfalcon, linuxppc-dev
During MTU change, the following events may happen.
Client-driven CRQ initialization fails due to partner’s CRQ closed,
causing client to enqueue a reset task for FATAL_ERROR. Then passive
(server-driven) CRQ initialization succeeds, causing client to
release CRQ and enqueue a reset task for failover. If the passive
CRQ initialization occurs before the FATAL reset task is processed,
the FATAL error reset task would try to access a CRQ message queue
that was freed, causing an oops. The problem may be most likely to
occur during DLPAR add vNIC with a non-default MTU, because the DLPAR
process will automatically issue a change MTU request.
Fix this by not processing fatal error reset if CRQ is passively
initialized after client-driven CRQ initialization fails.
Signed-off-by: Juliet Kim <julietk@linux.vnet.ibm.com>
---
drivers/net/ethernet/ibm/ibmvnic.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/ibm/ibmvnic.c b/drivers/net/ethernet/ibm/ibmvnic.c
index 4bd33245bad6..3de549c6c693 100644
--- a/drivers/net/ethernet/ibm/ibmvnic.c
+++ b/drivers/net/ethernet/ibm/ibmvnic.c
@@ -2189,7 +2189,8 @@ static void __ibmvnic_reset(struct work_struct *work)
rc = do_hard_reset(adapter, rwi, reset_state);
rtnl_unlock();
}
- } else {
+ } else if (!(rwi->reset_reason == VNIC_RESET_FATAL &&
+ adapter->from_passive_init)) {
rc = do_reset(adapter, rwi, reset_state);
}
kfree(rwi);
--
2.16.4
^ permalink raw reply related
* Re: [PATCH v2 2/3] mm/memory_hotplug: Introduce MHP_NO_FIRMWARE_MEMMAP
From: David Hildenbrand @ 2020-04-30 18:43 UTC (permalink / raw)
To: Eric W. Biederman, Andrew Morton
Cc: virtio-dev, linux-hyperv, Michal Hocko, Baoquan He, linux-mm,
Wei Yang, linux-s390, linux-nvdimm, linux-kernel, virtualization,
linux-acpi, Michael S . Tsirkin, Pankaj Gupta, xen-devel,
Michal Hocko, linuxppc-dev
In-Reply-To: <875zdg26hp.fsf@x220.int.ebiederm.org>
>>> If the class of memory is different then please by all means let's mark
>>> it differently in struct resource so everyone knows it is different.
>>> But that difference needs to be more than hotplug.
>>>
>>> That difference needs to be the hypervisor loaned us memory and might
>>> take it back at any time, or this memory is persistent and so it has
>>> these different characteristics so don't use it as ordinary ram.
>>
>> Yes, and I think kmem took an excellent approach of explicitly putting
>> that "System RAM" into a resource hierarchy. That "System RAM" won't
>> show up as a root node under /proc/iomem (see patch #3), which already
>> results in kexec-tools to treat it in a special way. I am thinking about
>> doing the same for virtio-mem.
>
> Reading this and your patch cover letters again my concern is that
> the justification seems to be letting the tail wag the dog.
>
> You want kexec-tools to behave in a certain way so you are changing the
> kernel.
>
> Rather it should be change the kernel to clearly reflect reality and if
> you can get away without a change to kexec-tools that is a bonus.
>
Right, because user space has to have a way to figure out what to do.
But talking about the firmware memmap, indicating something via a "raw
firmware-provided memory map", that is not actually in the "raw
firmware-provided memory map" feels wrong to me. (below)
>>> That information is also useful to other people looking at the system
>>> and seeing what is going on.
>>>
>>> Just please don't muddle the concepts, or assume that whatever subset of
>>> hotplug memory you are dealing with is the only subset.
>>
>> I can certainly rephrase the subject/description/comment, stating that
>> this is not to be used for ordinary hotplugged DIMMs - only when the
>> device driver is under control to decide what to do with that memory -
>> especially when kexec'ing.
>>
>> (previously, I called this flag MHP_DRIVER_MANAGED, but I think
>> MHP_NO_FIRMWARE_MEMMAP is clearer, we just need a better description)
>>
>> Would that make it clearer?
>
> I am not certain, but Andrew Morton deliberately added that
> firmware_map_add_hotplug call. Which means that there is a reason
> for putting hotplugged memory in the firmware map.
>
> So the justification needs to take that reason into account. The
> justification can not be it is hotplugged therefore it should not belong
> in the firmware memory map. Unless you can show that
> firmware_map_add_hotplug that was actually a bug and should be removed.
> But as it has been that way since 2010 that seems like a long shot.
>
> So my question is what is right for the firmware map?
We have documentation for that since 2008. Andrews patch is from 2010.
Documentation/ABI/testing/sysfs-firmware-memmap
It clearly talks about "raw firmware-provided memory map" and why the
interface was introduced at all ("on most architectures that
firmware-provided memory map is modified afterwards by the kernel itself").
>
> Why does the firmware map support hotplug entries?
I assume:
The firmware memmap was added primarily for x86-64 kexec (and still, is
mostly used on x86-64 only IIRC). There, we had ACPI hotplug. When DIMMs
get hotplugged on real HW, they get added to e820. Same applies to
memory added via HyperV balloon (unless memory is unplugged via
ballooning and you reboot ... the the e820 is changed as well). I assume
we wanted to be able to reflect that, to make kexec look like a real reboot.
This worked for a while. Then came dax/kmem. Now comes virtio-mem.
But I assume only Andrew can enlighten us.
@Andrew, any guidance here? Should we really add all memory to the
firmware memmap, even if this contradicts with the existing
documentation? (especially, if the actual firmware memmap will *not*
contain that memory after a reboot)
--
Thanks,
David / dhildenb
^ permalink raw reply
* Re: [PATCH v2 2/3] mm/memory_hotplug: Introduce MHP_NO_FIRMWARE_MEMMAP
From: Dan Williams @ 2020-04-30 18:58 UTC (permalink / raw)
To: David Hildenbrand
Cc: virtio-dev, linux-hyperv, Michal Hocko, Baoquan He, Linux ACPI,
Wei Yang, linux-s390, linux-nvdimm, Linux Kernel Mailing List,
virtualization, Linux MM, Michael S . Tsirkin, Eric W. Biederman,
Pankaj Gupta, xen-devel, Andrew Morton, Michal Hocko,
linuxppc-dev
In-Reply-To: <b28c9e02-8cf2-33ae-646b-fe50a185738e@redhat.com>
On Thu, Apr 30, 2020 at 11:44 AM David Hildenbrand <david@redhat.com> wrote:
>
> >>> If the class of memory is different then please by all means let's mark
> >>> it differently in struct resource so everyone knows it is different.
> >>> But that difference needs to be more than hotplug.
> >>>
> >>> That difference needs to be the hypervisor loaned us memory and might
> >>> take it back at any time, or this memory is persistent and so it has
> >>> these different characteristics so don't use it as ordinary ram.
> >>
> >> Yes, and I think kmem took an excellent approach of explicitly putting
> >> that "System RAM" into a resource hierarchy. That "System RAM" won't
> >> show up as a root node under /proc/iomem (see patch #3), which already
> >> results in kexec-tools to treat it in a special way. I am thinking about
> >> doing the same for virtio-mem.
> >
> > Reading this and your patch cover letters again my concern is that
> > the justification seems to be letting the tail wag the dog.
> >
> > You want kexec-tools to behave in a certain way so you are changing the
> > kernel.
> >
> > Rather it should be change the kernel to clearly reflect reality and if
> > you can get away without a change to kexec-tools that is a bonus.
> >
>
> Right, because user space has to have a way to figure out what to do.
>
> But talking about the firmware memmap, indicating something via a "raw
> firmware-provided memory map", that is not actually in the "raw
> firmware-provided memory map" feels wrong to me. (below)
>
>
> >>> That information is also useful to other people looking at the system
> >>> and seeing what is going on.
> >>>
> >>> Just please don't muddle the concepts, or assume that whatever subset of
> >>> hotplug memory you are dealing with is the only subset.
> >>
> >> I can certainly rephrase the subject/description/comment, stating that
> >> this is not to be used for ordinary hotplugged DIMMs - only when the
> >> device driver is under control to decide what to do with that memory -
> >> especially when kexec'ing.
> >>
> >> (previously, I called this flag MHP_DRIVER_MANAGED, but I think
> >> MHP_NO_FIRMWARE_MEMMAP is clearer, we just need a better description)
> >>
> >> Would that make it clearer?
> >
> > I am not certain, but Andrew Morton deliberately added that
> > firmware_map_add_hotplug call. Which means that there is a reason
> > for putting hotplugged memory in the firmware map.
> >
> > So the justification needs to take that reason into account. The
> > justification can not be it is hotplugged therefore it should not belong
> > in the firmware memory map. Unless you can show that
> > firmware_map_add_hotplug that was actually a bug and should be removed.
> > But as it has been that way since 2010 that seems like a long shot.
> >
> > So my question is what is right for the firmware map?
>
> We have documentation for that since 2008. Andrews patch is from 2010.
>
> Documentation/ABI/testing/sysfs-firmware-memmap
>
> It clearly talks about "raw firmware-provided memory map" and why the
> interface was introduced at all ("on most architectures that
> firmware-provided memory map is modified afterwards by the kernel itself").
>
> >
> > Why does the firmware map support hotplug entries?
>
> I assume:
>
> The firmware memmap was added primarily for x86-64 kexec (and still, is
> mostly used on x86-64 only IIRC). There, we had ACPI hotplug. When DIMMs
> get hotplugged on real HW, they get added to e820. Same applies to
> memory added via HyperV balloon (unless memory is unplugged via
> ballooning and you reboot ... the the e820 is changed as well). I assume
> we wanted to be able to reflect that, to make kexec look like a real reboot.
I can at least say that this breakdown makes sense to me. Traditional
memory hotplug results in permanent change to the raw firmware memory
map reported by the host at next reboot. These device-driver-owned
memory regions really want a hotplug policy per-kernel boot instance
and should fall back to the default reserved state at reboot (kexec or
otherwise). When I say hotplug-policy I mean whether the current
kernel wants to treat the device range as System RAM or leave it as
device-managed. The intent is that the follow-on kernel needs to
re-decide the device policy.
>
> This worked for a while. Then came dax/kmem. Now comes virtio-mem.
>
^ permalink raw reply
* [PATCH v3 0/2] PCI/ERR: Allow Native AER/DPC using _OSC
From: Jon Derrick @ 2020-04-30 18:46 UTC (permalink / raw)
To: Bjorn Helgaas
Cc: Kuppuswamy Sathyanarayanan, Rajat Jain, Frederick Lawler,
Sam Bobroff, linux-pci, Rafael J. Wysocki, linuxppc-dev,
linux-kernel, Olof Johansson, Alex Williamson, Patel, Mayurkumar,
Oliver O'Halloran, Bjorn Helgaas, Andy Shevchenko,
Mika Westerberg, Jon Derrick
Hi Bjorn & Kuppuswamy,
I see a problem in the DPC ECN [1] to _OSC in that it doesn't give us a way to
determine if firmware supports _OSC DPC negotation, and therefore how to handle
DPC.
Here is the wording of the ECN that implies that Firmware without _OSC DPC
negotiation support should have the OSPM rely on _OSC AER negotiation when
determining DPC control:
PCIe Base Specification suggests that Downstream Port Containment may be
controlled either by the Firmware or the Operating System. It also suggests
that the Firmware retain ownership of Downstream Port Containment if it also
owns AER. When the Firmware owns Downstream Port Containment, it is expected
to use the new "Error Disconnect Recover" notification to alert OSPM of a
Downstream Port Containment event.
In legacy platforms, as bits in _OSC are reserved prior to implementation, ACPI
Root Bus enumeration will mark these Host Bridges as without Native DPC
support, even though the specification implies it's expected that AER _OSC
negotiation determines DPC control for these platforms. There seems to be a
need for a way to determine if the DPC control bit in _OSC is supported and
fallback on AER otherwise.
Currently portdrv assumes DPC control if the port has Native AER services:
static int get_port_device_capability(struct pci_dev *dev)
...
if (pci_find_ext_capability(dev, PCI_EXT_CAP_ID_DPC) &&
pci_aer_available() &&
(pcie_ports_dpc_native || (services & PCIE_PORT_SERVICE_AER)))
services |= PCIE_PORT_SERVICE_DPC;
Newer firmware may not grant OSPM DPC control, if for instance, it expects to
use Error Disconnect Recovery. However it looks like ACPI will use DPC services
via the EDR driver, without binding the full DPC port service driver.
If we change portdrv to probe based on host->native_dpc and not AER, then we
break instances with legacy firmware where OSPM will clear host->native_dpc
solely due to _OSC bits being reserved:
struct pci_bus *acpi_pci_root_create(struct acpi_pci_root *root,
...
if (!(root->osc_control_set & OSC_PCI_EXPRESS_DPC_CONTROL))
host_bridge->native_dpc = 0;
So my assumption instead is that host->native_dpc can be 0 and expect Native
DPC services if AER is used. In other words, if and only if DPC probe is
invoked from portdrv, then it needs to rely on the AER dependency. Otherwise it
should be assumed that ACPI set up DPC via EDR. This covers legacy firmware.
However it seems like that could be trouble with newer firmware that might give
OSPM control of AER but not DPC, and would result in both Native DPC and EDR
being in effect.
Anyways here are two patches that give control of AER and DPC on the results of
_OSC. They don't mess with the HEST parser as I expect those to be removed at
some point. I need these for VMD support which doesn't even rely on _OSC, but I
suspect this won't be the last effort as we detangle Firmware First.
[1] https://members.pcisig.com/wg/PCI-SIG/document/12888
Jon Derrick (2):
PCI/AER: Use _OSC to determine Firmware First before HEST
PCI/DPC: Use _OSC to determine DPC support
drivers/pci/pcie/aer.c | 3 +++
drivers/pci/pcie/dpc.c | 3 ---
drivers/pci/pcie/portdrv_core.c | 3 ++-
3 files changed, 5 insertions(+), 4 deletions(-)
--
1.8.3.1
^ permalink raw reply
* [PATCH v3 1/2] PCI/AER: Use _OSC to determine Firmware First before HEST
From: Jon Derrick @ 2020-04-30 18:46 UTC (permalink / raw)
To: Bjorn Helgaas
Cc: Kuppuswamy Sathyanarayanan, Rajat Jain, Frederick Lawler,
Sam Bobroff, linux-pci, Rafael J. Wysocki, linuxppc-dev,
linux-kernel, Olof Johansson, Alex Williamson, Patel, Mayurkumar,
Oliver O'Halloran, Bjorn Helgaas, Andy Shevchenko,
Mika Westerberg, Jon Derrick
In-Reply-To: <1588272369-2145-1-git-send-email-jonathan.derrick@intel.com>
After a5bf8719af: "PCI/AER: Use only _OSC to determine AER ownership",
_OSC is the primary determiner of ownership of Firmware First error
handling rather than HEST.
ACPI Root Bus enumeration has been modified to flag Host Bridge devices
as using Native AER when _OSC has been negotiated for AER services.
This patch ensures the PCI layers first uses the _OSC negotiated state
by checking the Host Bridge's Native AER flag prior to HEST parsing.
Signed-off-by: Jon Derrick <jonathan.derrick@intel.com>
---
drivers/pci/pcie/aer.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/drivers/pci/pcie/aer.c b/drivers/pci/pcie/aer.c
index efc2677..f3d02f4 100644
--- a/drivers/pci/pcie/aer.c
+++ b/drivers/pci/pcie/aer.c
@@ -314,6 +314,9 @@ int pcie_aer_get_firmware_first(struct pci_dev *dev)
if (pcie_ports_native)
return 0;
+ if (pci_find_host_bridge(dev->bus)->native_aer)
+ return 0;
+
if (!dev->__aer_firmware_first_valid)
aer_set_firmware_first(dev);
return dev->__aer_firmware_first;
--
1.8.3.1
^ permalink raw reply related
* [PATCH v3 2/2] PCI/DPC: Use _OSC to determine DPC support
From: Jon Derrick @ 2020-04-30 18:46 UTC (permalink / raw)
To: Bjorn Helgaas
Cc: Kuppuswamy Sathyanarayanan, Rajat Jain, Frederick Lawler,
Sam Bobroff, linux-pci, Rafael J. Wysocki, linuxppc-dev,
linux-kernel, Olof Johansson, Alex Williamson, Patel, Mayurkumar,
Oliver O'Halloran, Bjorn Helgaas, Andy Shevchenko,
Mika Westerberg, Jon Derrick
In-Reply-To: <1588272369-2145-1-git-send-email-jonathan.derrick@intel.com>
After a5bf8719af: "PCI/AER: Use only _OSC to determine AER ownership",
_OSC is the primary determiner of ownership of Firmware First error
handling rather than HEST.
With the addition of DPC to _OSC [1], OSPM is able to negotiate DPC
services from Firmware. ACPI Root Bus enumeration sets the Host Bridge's
Native DPC flag on negotiation of those service. This patch changes DPC
probing to check DPC control as determined by _OSC, by checking the Host
Bridge's Native DPC member.
As most older platforms won't have DPC negotiable by _OSC, this patch
doesn't attempt to change behavior that assumes if OSPM has negotiated
AER by _OSC, OSPM will also want DPC control.
[1] https://members.pcisig.com/wg/PCI-SIG/document/12888
Signed-off-by: Jon Derrick <jonathan.derrick@intel.com>
---
drivers/pci/pcie/dpc.c | 3 ---
drivers/pci/pcie/portdrv_core.c | 3 ++-
2 files changed, 2 insertions(+), 4 deletions(-)
diff --git a/drivers/pci/pcie/dpc.c b/drivers/pci/pcie/dpc.c
index 7621704..9104929 100644
--- a/drivers/pci/pcie/dpc.c
+++ b/drivers/pci/pcie/dpc.c
@@ -284,9 +284,6 @@ static int dpc_probe(struct pcie_device *dev)
int status;
u16 ctl, cap;
- if (pcie_aer_get_firmware_first(pdev) && !pcie_ports_dpc_native)
- return -ENOTSUPP;
-
status = devm_request_threaded_irq(device, dev->irq, dpc_irq,
dpc_handler, IRQF_SHARED,
"pcie-dpc", pdev);
diff --git a/drivers/pci/pcie/portdrv_core.c b/drivers/pci/pcie/portdrv_core.c
index 50a9522..f2139a1 100644
--- a/drivers/pci/pcie/portdrv_core.c
+++ b/drivers/pci/pcie/portdrv_core.c
@@ -256,7 +256,8 @@ static int get_port_device_capability(struct pci_dev *dev)
*/
if (pci_find_ext_capability(dev, PCI_EXT_CAP_ID_DPC) &&
pci_aer_available() &&
- (pcie_ports_dpc_native || (services & PCIE_PORT_SERVICE_AER)))
+ (pcie_ports_dpc_native || host->native_dpc ||
+ (services & PCIE_PORT_SERVICE_AER)))
services |= PCIE_PORT_SERVICE_DPC;
if (pci_pcie_type(dev) == PCI_EXP_TYPE_DOWNSTREAM ||
--
1.8.3.1
^ permalink raw reply related
* Re: [PATCH 00/37] net: manually convert files to ReST format - part 2
From: David Miller @ 2020-04-30 19:56 UTC (permalink / raw)
To: mchehab+huawei
Cc: marcelo.leitner, linux-doc, dhowells, linux-sctp, paulus,
ioana.ciornei, linux-afs, devel, rds-devel, socketcan,
ruxandra.radulescu, corbet, linux-rdma, kadlec, coreteam, kuba,
pablo, linux-x25, santosh.shilimkar, linux-can, mkl, courmisch,
kvalo, fw, nhorman, geoff, netdev, vyasevich, linux-wireless,
linux-kernel, netfilter-devel, gregkh, johannes, linuxppc-dev
In-Reply-To: <cover.1588261997.git.mchehab+huawei@kernel.org>
From: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Date: Thu, 30 Apr 2020 18:03:55 +0200
> That's the second part of my work to convert the networking
> text files into ReST. it is based on today's linux-next (next-20200430).
>
> The full series (including those ones) are at:
>
> https://git.linuxtv.org/mchehab/experimental.git/log/?h=net-docs
>
> I should be sending the remaining patches (another /38 series)
> after getting those merged at -next.
>
> The documents, converted to HTML via the building system are at:
>
> https://www.infradead.org/~mchehab/kernel_docs/networking/
Series applied to net-next, thank you.
^ permalink raw reply
* Re: [PATCH net] ibmvnic: Skip fatal error reset after passive init
From: David Miller @ 2020-04-30 20:28 UTC (permalink / raw)
To: julietk; +Cc: netdev, linuxppc-dev, tlfalcon
In-Reply-To: <20200430182211.24211-1-julietk@linux.vnet.ibm.com>
From: Juliet Kim <julietk@linux.vnet.ibm.com>
Date: Thu, 30 Apr 2020 13:22:11 -0500
> During MTU change, the following events may happen.
> Client-driven CRQ initialization fails due to partner’s CRQ closed,
> causing client to enqueue a reset task for FATAL_ERROR. Then passive
> (server-driven) CRQ initialization succeeds, causing client to
> release CRQ and enqueue a reset task for failover. If the passive
> CRQ initialization occurs before the FATAL reset task is processed,
> the FATAL error reset task would try to access a CRQ message queue
> that was freed, causing an oops. The problem may be most likely to
> occur during DLPAR add vNIC with a non-default MTU, because the DLPAR
> process will automatically issue a change MTU request.
>
> Fix this by not processing fatal error reset if CRQ is passively
> initialized after client-driven CRQ initialization fails.
>
> Signed-off-by: Juliet Kim <julietk@linux.vnet.ibm.com>
Applied, thanks.
^ permalink raw reply
* [PATCH V1 00/10] Remove duplicated kmap code
From: ira.weiny @ 2020-04-30 20:38 UTC (permalink / raw)
To: linux-kernel, Andrew Morton, Christian Koenig, Huang Rui
Cc: Peter Zijlstra, Dave Hansen, dri-devel, James E.J. Bottomley,
Max Filippov, Paul Mackerras, H. Peter Anvin, sparclinux,
Ira Weiny, Thomas Gleixner, Helge Deller, x86, linux-csky,
Ingo Molnar, linux-snps-arc, linux-xtensa, Borislav Petkov,
Andy Lutomirski, Dan Williams, linux-arm-kernel, Chris Zankel,
Thomas Bogendoerfer, linux-parisc, linux-mips, linuxppc-dev,
David S. Miller
From: Ira Weiny <ira.weiny@intel.com>
The kmap infrastructure has been copied almost verbatim to every architecture.
This series consolidates obvious duplicated code by defining core functions
which call into the architectures only when needed.
Some of the k[un]map_atomic() implementations have some similarities but the
similarities were not sufficient to warrant further changes.
In addition we remove a duplicate implementation of kmap() in DRM.
Testing was done by 0day to cover all the architectures I can't readily
build/test.
---
Changes from V0:
rebase to 5.7-rc4
Define kmap_flush_tlb() and make kmap() truely arch independent.
Redefine the k[un]map_atomic_* code to call into the architectures for
high mem pages
Ensure all architectures define kmap_prot, use it appropriately, and
define kmap_atomic_prot()
Remove drm implementation of kmap_atomic()
Ira Weiny (10):
arch/kmap: Remove BUG_ON()
arch/xtensa: Move kmap build bug out of the way
arch/kmap: Remove redundant arch specific kmaps
arch/kunmap: Remove duplicate kunmap implementations
arch/kmap_atomic: Consolidate duplicate code
arch/kunmap_atomic: Consolidate duplicate code
arch/kmap: Ensure kmap_prot visibility
arch/kmap: Don't hard code kmap_prot values
arch/kmap: Define kmap_atomic_prot() for all arch's
drm: Remove drm specific kmap_atomic code
arch/arc/include/asm/highmem.h | 16 +------
arch/arc/mm/highmem.c | 28 +++----------
arch/arm/include/asm/highmem.h | 9 +---
arch/arm/mm/highmem.c | 35 +++-------------
arch/csky/include/asm/highmem.h | 11 ++---
arch/csky/mm/highmem.c | 43 +++++--------------
arch/microblaze/include/asm/highmem.h | 29 ++-----------
arch/microblaze/mm/highmem.c | 16 ++-----
arch/microblaze/mm/init.c | 3 --
arch/mips/include/asm/highmem.h | 11 ++---
arch/mips/mm/cache.c | 6 +--
arch/mips/mm/highmem.c | 49 ++++------------------
arch/nds32/include/asm/highmem.h | 9 +---
arch/nds32/mm/highmem.c | 39 +++--------------
arch/parisc/include/asm/cacheflush.h | 4 +-
arch/powerpc/include/asm/highmem.h | 30 ++------------
arch/powerpc/mm/highmem.c | 21 ++--------
arch/powerpc/mm/mem.c | 3 --
arch/sparc/include/asm/highmem.h | 23 +---------
arch/sparc/mm/highmem.c | 18 +++-----
arch/x86/include/asm/highmem.h | 11 +----
arch/x86/mm/highmem_32.c | 50 ++--------------------
arch/xtensa/include/asm/highmem.h | 28 +------------
arch/xtensa/mm/highmem.c | 23 +++++-----
drivers/gpu/drm/ttm/ttm_bo_util.c | 56 ++-----------------------
drivers/gpu/drm/vmwgfx/vmwgfx_blit.c | 16 +++----
include/drm/ttm/ttm_bo_api.h | 4 --
include/linux/highmem.h | 60 +++++++++++++++++++++++++--
28 files changed, 159 insertions(+), 492 deletions(-)
--
2.25.1
^ permalink raw reply
* [PATCH V1 01/10] arch/kmap: Remove BUG_ON()
From: ira.weiny @ 2020-04-30 20:38 UTC (permalink / raw)
To: linux-kernel, Andrew Morton, Christian Koenig, Huang Rui
Cc: Peter Zijlstra, Dave Hansen, dri-devel, James E.J. Bottomley,
Max Filippov, Paul Mackerras, H. Peter Anvin, sparclinux,
Ira Weiny, Dan Williams, Helge Deller, x86, linux-csky,
Ingo Molnar, linux-snps-arc, linux-xtensa, Borislav Petkov,
Andy Lutomirski, Thomas Gleixner, linux-arm-kernel, Chris Zankel,
Thomas Bogendoerfer, linux-parisc, linux-mips, linuxppc-dev,
David S. Miller
In-Reply-To: <20200430203845.582900-1-ira.weiny@intel.com>
From: Ira Weiny <ira.weiny@intel.com>
Replace the use of BUG_ON(in_interrupt()) in the kmap() and kunmap()
in favor of might_sleep().
Besides the benefits of might_sleep(), this normalizes the
implementations such that they can be made generic in subsequent
patches.
Reviewed-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: Ira Weiny <ira.weiny@intel.com>
---
arch/arc/include/asm/highmem.h | 2 +-
arch/arc/mm/highmem.c | 2 +-
arch/arm/mm/highmem.c | 2 +-
arch/csky/mm/highmem.c | 2 +-
arch/microblaze/include/asm/highmem.h | 2 +-
arch/mips/mm/highmem.c | 2 +-
arch/nds32/mm/highmem.c | 2 +-
arch/powerpc/include/asm/highmem.h | 2 +-
arch/sparc/include/asm/highmem.h | 4 ++--
arch/x86/mm/highmem_32.c | 3 +--
arch/xtensa/include/asm/highmem.h | 4 ++--
11 files changed, 13 insertions(+), 14 deletions(-)
diff --git a/arch/arc/include/asm/highmem.h b/arch/arc/include/asm/highmem.h
index 1af00accb37f..042e92921c4c 100644
--- a/arch/arc/include/asm/highmem.h
+++ b/arch/arc/include/asm/highmem.h
@@ -45,7 +45,7 @@ static inline void flush_cache_kmaps(void)
static inline void kunmap(struct page *page)
{
- BUG_ON(in_interrupt());
+ might_sleep();
if (!PageHighMem(page))
return;
kunmap_high(page);
diff --git a/arch/arc/mm/highmem.c b/arch/arc/mm/highmem.c
index fc8849e4f72e..39ef7b9a3aa9 100644
--- a/arch/arc/mm/highmem.c
+++ b/arch/arc/mm/highmem.c
@@ -51,7 +51,7 @@ static pte_t * fixmap_page_table;
void *kmap(struct page *page)
{
- BUG_ON(in_interrupt());
+ might_sleep();
if (!PageHighMem(page))
return page_address(page);
diff --git a/arch/arm/mm/highmem.c b/arch/arm/mm/highmem.c
index a76f8ace9ce6..cc6eb79ef20c 100644
--- a/arch/arm/mm/highmem.c
+++ b/arch/arm/mm/highmem.c
@@ -42,7 +42,7 @@ EXPORT_SYMBOL(kmap);
void kunmap(struct page *page)
{
- BUG_ON(in_interrupt());
+ might_sleep();
if (!PageHighMem(page))
return;
kunmap_high(page);
diff --git a/arch/csky/mm/highmem.c b/arch/csky/mm/highmem.c
index 813129145f3d..690d678649d1 100644
--- a/arch/csky/mm/highmem.c
+++ b/arch/csky/mm/highmem.c
@@ -29,7 +29,7 @@ EXPORT_SYMBOL(kmap);
void kunmap(struct page *page)
{
- BUG_ON(in_interrupt());
+ might_sleep();
if (!PageHighMem(page))
return;
kunmap_high(page);
diff --git a/arch/microblaze/include/asm/highmem.h b/arch/microblaze/include/asm/highmem.h
index 332c78e15198..99ced7278b5c 100644
--- a/arch/microblaze/include/asm/highmem.h
+++ b/arch/microblaze/include/asm/highmem.h
@@ -66,7 +66,7 @@ static inline void *kmap(struct page *page)
static inline void kunmap(struct page *page)
{
- BUG_ON(in_interrupt());
+ might_sleep();
if (!PageHighMem(page))
return;
kunmap_high(page);
diff --git a/arch/mips/mm/highmem.c b/arch/mips/mm/highmem.c
index d08e6d7d533b..edd889f6cede 100644
--- a/arch/mips/mm/highmem.c
+++ b/arch/mips/mm/highmem.c
@@ -28,7 +28,7 @@ EXPORT_SYMBOL(kmap);
void kunmap(struct page *page)
{
- BUG_ON(in_interrupt());
+ might_sleep();
if (!PageHighMem(page))
return;
kunmap_high(page);
diff --git a/arch/nds32/mm/highmem.c b/arch/nds32/mm/highmem.c
index 022779af6148..4c7c28e994ea 100644
--- a/arch/nds32/mm/highmem.c
+++ b/arch/nds32/mm/highmem.c
@@ -24,7 +24,7 @@ EXPORT_SYMBOL(kmap);
void kunmap(struct page *page)
{
- BUG_ON(in_interrupt());
+ might_sleep();
if (!PageHighMem(page))
return;
kunmap_high(page);
diff --git a/arch/powerpc/include/asm/highmem.h b/arch/powerpc/include/asm/highmem.h
index a4b65b186ec6..529512f6d65a 100644
--- a/arch/powerpc/include/asm/highmem.h
+++ b/arch/powerpc/include/asm/highmem.h
@@ -74,7 +74,7 @@ static inline void *kmap(struct page *page)
static inline void kunmap(struct page *page)
{
- BUG_ON(in_interrupt());
+ might_sleep();
if (!PageHighMem(page))
return;
kunmap_high(page);
diff --git a/arch/sparc/include/asm/highmem.h b/arch/sparc/include/asm/highmem.h
index 18d776925c45..7dd2d4b3f980 100644
--- a/arch/sparc/include/asm/highmem.h
+++ b/arch/sparc/include/asm/highmem.h
@@ -55,7 +55,7 @@ void kunmap_high(struct page *page);
static inline void *kmap(struct page *page)
{
- BUG_ON(in_interrupt());
+ might_sleep();
if (!PageHighMem(page))
return page_address(page);
return kmap_high(page);
@@ -63,7 +63,7 @@ static inline void *kmap(struct page *page)
static inline void kunmap(struct page *page)
{
- BUG_ON(in_interrupt());
+ might_sleep();
if (!PageHighMem(page))
return;
kunmap_high(page);
diff --git a/arch/x86/mm/highmem_32.c b/arch/x86/mm/highmem_32.c
index 0a1898b8552e..8af66382672b 100644
--- a/arch/x86/mm/highmem_32.c
+++ b/arch/x86/mm/highmem_32.c
@@ -15,8 +15,7 @@ EXPORT_SYMBOL(kmap);
void kunmap(struct page *page)
{
- if (in_interrupt())
- BUG();
+ might_sleep();
if (!PageHighMem(page))
return;
kunmap_high(page);
diff --git a/arch/xtensa/include/asm/highmem.h b/arch/xtensa/include/asm/highmem.h
index 04e9340eac4b..413848cc1e56 100644
--- a/arch/xtensa/include/asm/highmem.h
+++ b/arch/xtensa/include/asm/highmem.h
@@ -73,7 +73,7 @@ static inline void *kmap(struct page *page)
*/
BUILD_BUG_ON(PKMAP_BASE <
TLBTEMP_BASE_1 + TLBTEMP_SIZE);
- BUG_ON(in_interrupt());
+ might_sleep();
if (!PageHighMem(page))
return page_address(page);
return kmap_high(page);
@@ -81,7 +81,7 @@ static inline void *kmap(struct page *page)
static inline void kunmap(struct page *page)
{
- BUG_ON(in_interrupt());
+ might_sleep();
if (!PageHighMem(page))
return;
kunmap_high(page);
--
2.25.1
^ permalink raw reply related
* [PATCH V1 02/10] arch/xtensa: Move kmap build bug out of the way
From: ira.weiny @ 2020-04-30 20:38 UTC (permalink / raw)
To: linux-kernel, Andrew Morton, Christian Koenig, Huang Rui
Cc: Peter Zijlstra, Dave Hansen, dri-devel, James E.J. Bottomley,
Max Filippov, Paul Mackerras, H. Peter Anvin, sparclinux,
Ira Weiny, Thomas Gleixner, Helge Deller, x86, linux-csky,
Ingo Molnar, linux-snps-arc, linux-xtensa, Borislav Petkov,
Andy Lutomirski, Dan Williams, linux-arm-kernel, Chris Zankel,
Thomas Bogendoerfer, linux-parisc, linux-mips, linuxppc-dev,
David S. Miller
In-Reply-To: <20200430203845.582900-1-ira.weiny@intel.com>
From: Ira Weiny <ira.weiny@intel.com>
Move the kmap() build bug to kmap_init() to facilitate patches to lift
kmap() to the core.
Signed-off-by: Ira Weiny <ira.weiny@intel.com>
---
arch/xtensa/include/asm/highmem.h | 5 -----
arch/xtensa/mm/highmem.c | 5 +++++
2 files changed, 5 insertions(+), 5 deletions(-)
diff --git a/arch/xtensa/include/asm/highmem.h b/arch/xtensa/include/asm/highmem.h
index 413848cc1e56..a9587c85be85 100644
--- a/arch/xtensa/include/asm/highmem.h
+++ b/arch/xtensa/include/asm/highmem.h
@@ -68,11 +68,6 @@ void kunmap_high(struct page *page);
static inline void *kmap(struct page *page)
{
- /* Check if this memory layout is broken because PKMAP overlaps
- * page table.
- */
- BUILD_BUG_ON(PKMAP_BASE <
- TLBTEMP_BASE_1 + TLBTEMP_SIZE);
might_sleep();
if (!PageHighMem(page))
return page_address(page);
diff --git a/arch/xtensa/mm/highmem.c b/arch/xtensa/mm/highmem.c
index 184ceadccc1a..711641c4d214 100644
--- a/arch/xtensa/mm/highmem.c
+++ b/arch/xtensa/mm/highmem.c
@@ -88,6 +88,11 @@ void __init kmap_init(void)
{
unsigned long kmap_vstart;
+ /* Check if this memory layout is broken because PKMAP overlaps
+ * page table.
+ */
+ BUILD_BUG_ON(PKMAP_BASE <
+ TLBTEMP_BASE_1 + TLBTEMP_SIZE);
/* cache the first kmap pte */
kmap_vstart = __fix_to_virt(FIX_KMAP_BEGIN);
kmap_pte = kmap_get_fixmap_pte(kmap_vstart);
--
2.25.1
^ permalink raw reply related
* [PATCH V1 05/10] arch/kmap_atomic: Consolidate duplicate code
From: ira.weiny @ 2020-04-30 20:38 UTC (permalink / raw)
To: linux-kernel, Andrew Morton, Christian Koenig, Huang Rui
Cc: Peter Zijlstra, Dave Hansen, dri-devel, James E.J. Bottomley,
Max Filippov, Paul Mackerras, H. Peter Anvin, sparclinux,
Ira Weiny, Thomas Gleixner, Helge Deller, x86, linux-csky,
Ingo Molnar, linux-snps-arc, linux-xtensa, Borislav Petkov,
Andy Lutomirski, Dan Williams, linux-arm-kernel, Chris Zankel,
Thomas Bogendoerfer, linux-parisc, linux-mips, linuxppc-dev,
David S. Miller
In-Reply-To: <20200430203845.582900-1-ira.weiny@intel.com>
From: Ira Weiny <ira.weiny@intel.com>
Every arch has the same code to ensure atomic operations and a check for
!HIGHMEM page.
Remove the duplicate code by defining a core kmap_atomic() which only
calls the arch specific kmap_atomic_high() when the page is high memory.
Signed-off-by: Ira Weiny <ira.weiny@intel.com>
---
Changes from V0:
consolidate comments
Use a similar architecture to kmap() and define
kmap_atomic_high() for architecture specific
functionality
Fix 0-day build issue in arch/mips/mm/cache.c
---
arch/arc/include/asm/highmem.h | 2 +-
arch/arc/mm/highmem.c | 9 ++-------
arch/arm/include/asm/highmem.h | 2 +-
arch/arm/mm/highmem.c | 9 ++-------
arch/csky/include/asm/highmem.h | 2 +-
arch/csky/mm/highmem.c | 9 ++-------
arch/microblaze/include/asm/highmem.h | 2 +-
arch/microblaze/mm/highmem.c | 6 ------
arch/mips/include/asm/highmem.h | 2 +-
arch/mips/mm/cache.c | 2 +-
arch/mips/mm/highmem.c | 18 ++----------------
arch/nds32/include/asm/highmem.h | 2 +-
arch/nds32/mm/highmem.c | 9 ++-------
arch/powerpc/include/asm/highmem.h | 2 +-
arch/powerpc/mm/highmem.c | 11 -----------
arch/sparc/include/asm/highmem.h | 2 +-
arch/sparc/mm/highmem.c | 9 ++-------
arch/x86/include/asm/highmem.h | 7 +++++--
arch/x86/mm/highmem_32.c | 20 --------------------
arch/xtensa/include/asm/highmem.h | 2 +-
arch/xtensa/mm/highmem.c | 9 ++-------
include/linux/highmem.h | 22 ++++++++++++++++++++++
22 files changed, 51 insertions(+), 107 deletions(-)
diff --git a/arch/arc/include/asm/highmem.h b/arch/arc/include/asm/highmem.h
index 8387a5596a91..75bd0fa77fe2 100644
--- a/arch/arc/include/asm/highmem.h
+++ b/arch/arc/include/asm/highmem.h
@@ -30,7 +30,7 @@
#include <asm/cacheflush.h>
-extern void *kmap_atomic(struct page *page);
+extern void *kmap_atomic_high(struct page *page);
extern void __kunmap_atomic(void *kvaddr);
extern void kmap_init(void);
diff --git a/arch/arc/mm/highmem.c b/arch/arc/mm/highmem.c
index 4db13a6b9f3b..0964b011c29f 100644
--- a/arch/arc/mm/highmem.c
+++ b/arch/arc/mm/highmem.c
@@ -49,16 +49,11 @@
extern pte_t * pkmap_page_table;
static pte_t * fixmap_page_table;
-void *kmap_atomic(struct page *page)
+void *kmap_atomic_high(struct page *page)
{
int idx, cpu_idx;
unsigned long vaddr;
- preempt_disable();
- pagefault_disable();
- if (!PageHighMem(page))
- return page_address(page);
-
cpu_idx = kmap_atomic_idx_push();
idx = cpu_idx + KM_TYPE_NR * smp_processor_id();
vaddr = FIXMAP_ADDR(idx);
@@ -68,7 +63,7 @@ void *kmap_atomic(struct page *page)
return (void *)vaddr;
}
-EXPORT_SYMBOL(kmap_atomic);
+EXPORT_SYMBOL(kmap_atomic_high);
void __kunmap_atomic(void *kv)
{
diff --git a/arch/arm/include/asm/highmem.h b/arch/arm/include/asm/highmem.h
index 736f65283e7b..4edb6db3a5c8 100644
--- a/arch/arm/include/asm/highmem.h
+++ b/arch/arm/include/asm/highmem.h
@@ -60,7 +60,7 @@ static inline void *kmap_high_get(struct page *page)
* when CONFIG_HIGHMEM is not set.
*/
#ifdef CONFIG_HIGHMEM
-extern void *kmap_atomic(struct page *page);
+extern void *kmap_atomic_high(struct page *page);
extern void __kunmap_atomic(void *kvaddr);
extern void *kmap_atomic_pfn(unsigned long pfn);
#endif
diff --git a/arch/arm/mm/highmem.c b/arch/arm/mm/highmem.c
index c700b32350ee..075fdc235091 100644
--- a/arch/arm/mm/highmem.c
+++ b/arch/arm/mm/highmem.c
@@ -31,18 +31,13 @@ static inline pte_t get_fixmap_pte(unsigned long vaddr)
return *ptep;
}
-void *kmap_atomic(struct page *page)
+void *kmap_atomic_high(struct page *page)
{
unsigned int idx;
unsigned long vaddr;
void *kmap;
int type;
- preempt_disable();
- pagefault_disable();
- if (!PageHighMem(page))
- return page_address(page);
-
#ifdef CONFIG_DEBUG_HIGHMEM
/*
* There is no cache coherency issue when non VIVT, so force the
@@ -76,7 +71,7 @@ void *kmap_atomic(struct page *page)
return (void *)vaddr;
}
-EXPORT_SYMBOL(kmap_atomic);
+EXPORT_SYMBOL(kmap_atomic_high);
void __kunmap_atomic(void *kvaddr)
{
diff --git a/arch/csky/include/asm/highmem.h b/arch/csky/include/asm/highmem.h
index be11c5b67122..6807df1232f3 100644
--- a/arch/csky/include/asm/highmem.h
+++ b/arch/csky/include/asm/highmem.h
@@ -32,7 +32,7 @@ extern pte_t *pkmap_page_table;
#define ARCH_HAS_KMAP_FLUSH_TLB
extern void kmap_flush_tlb(unsigned long addr);
-extern void *kmap_atomic(struct page *page);
+extern void *kmap_atomic_high(struct page *page);
extern void __kunmap_atomic(void *kvaddr);
extern void *kmap_atomic_pfn(unsigned long pfn);
extern struct page *kmap_atomic_to_page(void *ptr);
diff --git a/arch/csky/mm/highmem.c b/arch/csky/mm/highmem.c
index e9952211264b..63d74b47eee6 100644
--- a/arch/csky/mm/highmem.c
+++ b/arch/csky/mm/highmem.c
@@ -21,16 +21,11 @@ EXPORT_SYMBOL(kmap_flush_tlb);
EXPORT_SYMBOL(kmap);
-void *kmap_atomic(struct page *page)
+void *kmap_atomic_high(struct page *page)
{
unsigned long vaddr;
int idx, type;
- preempt_disable();
- pagefault_disable();
- if (!PageHighMem(page))
- return page_address(page);
-
type = kmap_atomic_idx_push();
idx = type + KM_TYPE_NR*smp_processor_id();
vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
@@ -42,7 +37,7 @@ void *kmap_atomic(struct page *page)
return (void *)vaddr;
}
-EXPORT_SYMBOL(kmap_atomic);
+EXPORT_SYMBOL(kmap_atomic_high);
void __kunmap_atomic(void *kvaddr)
{
diff --git a/arch/microblaze/include/asm/highmem.h b/arch/microblaze/include/asm/highmem.h
index 0c94046f2d58..fe4ad8bac9ae 100644
--- a/arch/microblaze/include/asm/highmem.h
+++ b/arch/microblaze/include/asm/highmem.h
@@ -54,7 +54,7 @@ extern pte_t *pkmap_page_table;
extern void *kmap_atomic_prot(struct page *page, pgprot_t prot);
extern void __kunmap_atomic(void *kvaddr);
-static inline void *kmap_atomic(struct page *page)
+static inline void *kmap_atomic_high(struct page *page)
{
return kmap_atomic_prot(page, kmap_prot);
}
diff --git a/arch/microblaze/mm/highmem.c b/arch/microblaze/mm/highmem.c
index d7569f77fa15..a14f356b055b 100644
--- a/arch/microblaze/mm/highmem.c
+++ b/arch/microblaze/mm/highmem.c
@@ -38,12 +38,6 @@ void *kmap_atomic_prot(struct page *page, pgprot_t prot)
unsigned long vaddr;
int idx, type;
- preempt_disable();
- pagefault_disable();
- if (!PageHighMem(page))
- return page_address(page);
-
-
type = kmap_atomic_idx_push();
idx = type + KM_TYPE_NR*smp_processor_id();
vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
diff --git a/arch/mips/include/asm/highmem.h b/arch/mips/include/asm/highmem.h
index 24e7e7e5cc7b..a515bcf15d4b 100644
--- a/arch/mips/include/asm/highmem.h
+++ b/arch/mips/include/asm/highmem.h
@@ -48,7 +48,7 @@ extern pte_t *pkmap_page_table;
#define ARCH_HAS_KMAP_FLUSH_TLB
extern void kmap_flush_tlb(unsigned long addr);
-extern void *kmap_atomic(struct page *page);
+extern void *kmap_atomic_high(struct page *page);
extern void __kunmap_atomic(void *kvaddr);
extern void *kmap_atomic_pfn(unsigned long pfn);
diff --git a/arch/mips/mm/cache.c b/arch/mips/mm/cache.c
index 33b409391ddb..f015bb51fab0 100644
--- a/arch/mips/mm/cache.c
+++ b/arch/mips/mm/cache.c
@@ -14,9 +14,9 @@
#include <linux/sched.h>
#include <linux/syscalls.h>
#include <linux/mm.h>
+#include <linux/highmem.h>
#include <asm/cacheflush.h>
-#include <asm/highmem.h>
#include <asm/processor.h>
#include <asm/cpu.h>
#include <asm/cpu-features.h>
diff --git a/arch/mips/mm/highmem.c b/arch/mips/mm/highmem.c
index eb8ec8493f2f..2bda56372995 100644
--- a/arch/mips/mm/highmem.c
+++ b/arch/mips/mm/highmem.c
@@ -18,25 +18,11 @@ void kmap_flush_tlb(unsigned long addr)
}
EXPORT_SYMBOL(kmap_flush_tlb);
-/*
- * kmap_atomic/kunmap_atomic is significantly faster than kmap/kunmap because
- * no global lock is needed and because the kmap code must perform a global TLB
- * invalidation when the kmap pool wraps.
- *
- * However when holding an atomic kmap is is not legal to sleep, so atomic
- * kmaps are appropriate for short, tight code paths only.
- */
-
-void *kmap_atomic(struct page *page)
+void *kmap_atomic_high(struct page *page)
{
unsigned long vaddr;
int idx, type;
- preempt_disable();
- pagefault_disable();
- if (!PageHighMem(page))
- return page_address(page);
-
type = kmap_atomic_idx_push();
idx = type + KM_TYPE_NR*smp_processor_id();
vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
@@ -48,7 +34,7 @@ void *kmap_atomic(struct page *page)
return (void*) vaddr;
}
-EXPORT_SYMBOL(kmap_atomic);
+EXPORT_SYMBOL(kmap_atomic_high);
void __kunmap_atomic(void *kvaddr)
{
diff --git a/arch/nds32/include/asm/highmem.h b/arch/nds32/include/asm/highmem.h
index c93c7368bb3f..28f5e7072c70 100644
--- a/arch/nds32/include/asm/highmem.h
+++ b/arch/nds32/include/asm/highmem.h
@@ -51,7 +51,7 @@ extern void kmap_init(void);
* when CONFIG_HIGHMEM is not set.
*/
#ifdef CONFIG_HIGHMEM
-extern void *kmap_atomic(struct page *page);
+extern void *kmap_atomic_high(struct page *page);
extern void __kunmap_atomic(void *kvaddr);
extern void *kmap_atomic_pfn(unsigned long pfn);
extern struct page *kmap_atomic_to_page(void *ptr);
diff --git a/arch/nds32/mm/highmem.c b/arch/nds32/mm/highmem.c
index f9348bec0ecb..f5f3a21460c4 100644
--- a/arch/nds32/mm/highmem.c
+++ b/arch/nds32/mm/highmem.c
@@ -10,18 +10,13 @@
#include <asm/fixmap.h>
#include <asm/tlbflush.h>
-void *kmap_atomic(struct page *page)
+void *kmap_atomic_high(struct page *page)
{
unsigned int idx;
unsigned long vaddr, pte;
int type;
pte_t *ptep;
- preempt_disable();
- pagefault_disable();
- if (!PageHighMem(page))
- return page_address(page);
-
type = kmap_atomic_idx_push();
idx = type + KM_TYPE_NR * smp_processor_id();
@@ -37,7 +32,7 @@ void *kmap_atomic(struct page *page)
return (void *)vaddr;
}
-EXPORT_SYMBOL(kmap_atomic);
+EXPORT_SYMBOL(kmap_atomic_high);
void __kunmap_atomic(void *kvaddr)
{
diff --git a/arch/powerpc/include/asm/highmem.h b/arch/powerpc/include/asm/highmem.h
index ba3371977d49..ac0efc2cf08a 100644
--- a/arch/powerpc/include/asm/highmem.h
+++ b/arch/powerpc/include/asm/highmem.h
@@ -62,7 +62,7 @@ extern pte_t *pkmap_page_table;
extern void *kmap_atomic_prot(struct page *page, pgprot_t prot);
extern void __kunmap_atomic(void *kvaddr);
-static inline void *kmap_atomic(struct page *page)
+static inline void *kmap_atomic_high(struct page *page)
{
return kmap_atomic_prot(page, kmap_prot);
}
diff --git a/arch/powerpc/mm/highmem.c b/arch/powerpc/mm/highmem.c
index 320c1672b2ae..f9558ef4b8fa 100644
--- a/arch/powerpc/mm/highmem.c
+++ b/arch/powerpc/mm/highmem.c
@@ -24,22 +24,11 @@
#include <linux/highmem.h>
#include <linux/module.h>
-/*
- * The use of kmap_atomic/kunmap_atomic is discouraged - kmap/kunmap
- * gives a more generic (and caching) interface. But kmap_atomic can
- * be used in IRQ contexts, so in some (very limited) cases we need
- * it.
- */
void *kmap_atomic_prot(struct page *page, pgprot_t prot)
{
unsigned long vaddr;
int idx, type;
- preempt_disable();
- pagefault_disable();
- if (!PageHighMem(page))
- return page_address(page);
-
type = kmap_atomic_idx_push();
idx = type + KM_TYPE_NR*smp_processor_id();
vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
diff --git a/arch/sparc/include/asm/highmem.h b/arch/sparc/include/asm/highmem.h
index 4bdb79fed02c..c96a0603d821 100644
--- a/arch/sparc/include/asm/highmem.h
+++ b/arch/sparc/include/asm/highmem.h
@@ -50,7 +50,7 @@ void kmap_init(void) __init;
#define PKMAP_END (PKMAP_ADDR(LAST_PKMAP))
-void *kmap_atomic(struct page *page);
+void *kmap_atomic_high(struct page *page);
void __kunmap_atomic(void *kvaddr);
#define flush_cache_kmaps() flush_cache_all()
diff --git a/arch/sparc/mm/highmem.c b/arch/sparc/mm/highmem.c
index d4a80adea7e5..b53070ab6a31 100644
--- a/arch/sparc/mm/highmem.c
+++ b/arch/sparc/mm/highmem.c
@@ -53,16 +53,11 @@ void __init kmap_init(void)
kmap_prot = __pgprot(SRMMU_ET_PTE | SRMMU_PRIV | SRMMU_CACHE);
}
-void *kmap_atomic(struct page *page)
+void *kmap_atomic_high(struct page *page)
{
unsigned long vaddr;
long idx, type;
- preempt_disable();
- pagefault_disable();
- if (!PageHighMem(page))
- return page_address(page);
-
type = kmap_atomic_idx_push();
idx = type + KM_TYPE_NR*smp_processor_id();
vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
@@ -87,7 +82,7 @@ void *kmap_atomic(struct page *page)
return (void*) vaddr;
}
-EXPORT_SYMBOL(kmap_atomic);
+EXPORT_SYMBOL(kmap_atomic_high);
void __kunmap_atomic(void *kvaddr)
{
diff --git a/arch/x86/include/asm/highmem.h b/arch/x86/include/asm/highmem.h
index 90b96594d6c5..72e154e17416 100644
--- a/arch/x86/include/asm/highmem.h
+++ b/arch/x86/include/asm/highmem.h
@@ -58,8 +58,11 @@ extern unsigned long highstart_pfn, highend_pfn;
#define PKMAP_NR(virt) ((virt-PKMAP_BASE) >> PAGE_SHIFT)
#define PKMAP_ADDR(nr) (PKMAP_BASE + ((nr) << PAGE_SHIFT))
-void *kmap_atomic_prot(struct page *page, pgprot_t prot);
-void *kmap_atomic(struct page *page);
+extern void *kmap_atomic_prot(struct page *page, pgprot_t prot);
+void *kmap_atomic_high(struct page *page)
+{
+ return kmap_atomic_prot(page, kmap_prot);
+}
void __kunmap_atomic(void *kvaddr);
void *kmap_atomic_pfn(unsigned long pfn);
void *kmap_atomic_prot_pfn(unsigned long pfn, pgprot_t prot);
diff --git a/arch/x86/mm/highmem_32.c b/arch/x86/mm/highmem_32.c
index c4ebfd0ae401..937d2cc40389 100644
--- a/arch/x86/mm/highmem_32.c
+++ b/arch/x86/mm/highmem_32.c
@@ -4,25 +4,11 @@
#include <linux/swap.h> /* for totalram_pages */
#include <linux/memblock.h>
-/*
- * kmap_atomic/kunmap_atomic is significantly faster than kmap/kunmap because
- * no global lock is needed and because the kmap code must perform a global TLB
- * invalidation when the kmap pool wraps.
- *
- * However when holding an atomic kmap it is not legal to sleep, so atomic
- * kmaps are appropriate for short, tight code paths only.
- */
void *kmap_atomic_prot(struct page *page, pgprot_t prot)
{
unsigned long vaddr;
int idx, type;
- preempt_disable();
- pagefault_disable();
-
- if (!PageHighMem(page))
- return page_address(page);
-
type = kmap_atomic_idx_push();
idx = type + KM_TYPE_NR*smp_processor_id();
vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
@@ -34,12 +20,6 @@ void *kmap_atomic_prot(struct page *page, pgprot_t prot)
}
EXPORT_SYMBOL(kmap_atomic_prot);
-void *kmap_atomic(struct page *page)
-{
- return kmap_atomic_prot(page, kmap_prot);
-}
-EXPORT_SYMBOL(kmap_atomic);
-
/*
* This is the same as kmap_atomic() but can map memory that doesn't
* have a struct page associated with it.
diff --git a/arch/xtensa/include/asm/highmem.h b/arch/xtensa/include/asm/highmem.h
index 5a481f7def0b..d4fb9f78ba32 100644
--- a/arch/xtensa/include/asm/highmem.h
+++ b/arch/xtensa/include/asm/highmem.h
@@ -68,7 +68,7 @@ static inline void flush_cache_kmaps(void)
flush_cache_all();
}
-void *kmap_atomic(struct page *page);
+void *kmap_atomic_high(struct page *page);
void __kunmap_atomic(void *kvaddr);
void kmap_init(void);
diff --git a/arch/xtensa/mm/highmem.c b/arch/xtensa/mm/highmem.c
index 711641c4d214..217f2ebaa298 100644
--- a/arch/xtensa/mm/highmem.c
+++ b/arch/xtensa/mm/highmem.c
@@ -37,16 +37,11 @@ static inline enum fixed_addresses kmap_idx(int type, unsigned long color)
color;
}
-void *kmap_atomic(struct page *page)
+void *kmap_atomic_high(struct page *page)
{
enum fixed_addresses idx;
unsigned long vaddr;
- preempt_disable();
- pagefault_disable();
- if (!PageHighMem(page))
- return page_address(page);
-
idx = kmap_idx(kmap_atomic_idx_push(),
DCACHE_ALIAS(page_to_phys(page)));
vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
@@ -57,7 +52,7 @@ void *kmap_atomic(struct page *page)
return (void *)vaddr;
}
-EXPORT_SYMBOL(kmap_atomic);
+EXPORT_SYMBOL(kmap_atomic_high);
void __kunmap_atomic(void *kvaddr)
{
diff --git a/include/linux/highmem.h b/include/linux/highmem.h
index ae6e8cb81043..e0106b4f7dbb 100644
--- a/include/linux/highmem.h
+++ b/include/linux/highmem.h
@@ -61,6 +61,28 @@ static inline void kunmap(struct page *page)
kunmap_high(page);
}
+/*
+ * kmap_atomic/kunmap_atomic is significantly faster than kmap/kunmap because
+ * no global lock is needed and because the kmap code must perform a global TLB
+ * invalidation when the kmap pool wraps.
+ *
+ * However when holding an atomic kmap is is not legal to sleep, so atomic
+ * kmaps are appropriate for short, tight code paths only.
+ *
+ * The use of kmap_atomic/kunmap_atomic is discouraged - kmap/kunmap
+ * gives a more generic (and caching) interface. But kmap_atomic can
+ * be used in IRQ contexts, so in some (very limited) cases we need
+ * it.
+ */
+static inline void *kmap_atomic(struct page *page)
+{
+ preempt_disable();
+ pagefault_disable();
+ if (!PageHighMem(page))
+ return page_address(page);
+ return kmap_atomic_high(page);
+}
+
/* declarations for linux/mm/highmem.c */
unsigned int nr_free_highpages(void);
extern atomic_long_t _totalhigh_pages;
--
2.25.1
^ permalink raw reply related
* [PATCH V1 04/10] arch/kunmap: Remove duplicate kunmap implementations
From: ira.weiny @ 2020-04-30 20:38 UTC (permalink / raw)
To: linux-kernel, Andrew Morton, Christian Koenig, Huang Rui
Cc: Peter Zijlstra, Dave Hansen, dri-devel, James E.J. Bottomley,
Max Filippov, Paul Mackerras, H. Peter Anvin, sparclinux,
Ira Weiny, Thomas Gleixner, Helge Deller, x86, linux-csky,
Ingo Molnar, linux-snps-arc, linux-xtensa, Borislav Petkov,
Andy Lutomirski, Dan Williams, linux-arm-kernel, Chris Zankel,
Thomas Bogendoerfer, linux-parisc, linux-mips, linuxppc-dev,
David S. Miller
In-Reply-To: <20200430203845.582900-1-ira.weiny@intel.com>
From: Ira Weiny <ira.weiny@intel.com>
All architectures do exactly the same thing for kunmap(); remove all the
duplicate definitions and lift the call to the core.
This also has the benefit of changing kmap_unmap() on a number of
architectures to be an inline call rather than an actual function.
Signed-off-by: Ira Weiny <ira.weiny@intel.com>
---
Changes from V0:
lift kunmap_high() declaration as well.
---
arch/arc/include/asm/highmem.h | 10 ----------
arch/arm/include/asm/highmem.h | 3 ---
arch/arm/mm/highmem.c | 9 ---------
arch/csky/include/asm/highmem.h | 3 ---
arch/csky/mm/highmem.c | 9 ---------
arch/microblaze/include/asm/highmem.h | 9 ---------
arch/mips/include/asm/highmem.h | 3 ---
arch/mips/mm/highmem.c | 9 ---------
arch/nds32/include/asm/highmem.h | 3 ---
arch/nds32/mm/highmem.c | 10 ----------
arch/powerpc/include/asm/highmem.h | 9 ---------
arch/sparc/include/asm/highmem.h | 10 ----------
arch/x86/include/asm/highmem.h | 4 ----
arch/x86/mm/highmem_32.c | 9 ---------
arch/xtensa/include/asm/highmem.h | 10 ----------
include/linux/highmem.h | 9 +++++++++
16 files changed, 9 insertions(+), 110 deletions(-)
diff --git a/arch/arc/include/asm/highmem.h b/arch/arc/include/asm/highmem.h
index 96eb67c86961..8387a5596a91 100644
--- a/arch/arc/include/asm/highmem.h
+++ b/arch/arc/include/asm/highmem.h
@@ -32,7 +32,6 @@
extern void *kmap_atomic(struct page *page);
extern void __kunmap_atomic(void *kvaddr);
-extern void kunmap_high(struct page *page);
extern void kmap_init(void);
@@ -41,15 +40,6 @@ static inline void flush_cache_kmaps(void)
flush_cache_all();
}
-static inline void kunmap(struct page *page)
-{
- might_sleep();
- if (!PageHighMem(page))
- return;
- kunmap_high(page);
-}
-
-
#endif
#endif
diff --git a/arch/arm/include/asm/highmem.h b/arch/arm/include/asm/highmem.h
index c917522541de..736f65283e7b 100644
--- a/arch/arm/include/asm/highmem.h
+++ b/arch/arm/include/asm/highmem.h
@@ -20,8 +20,6 @@
extern pte_t *pkmap_page_table;
-extern void kunmap_high(struct page *page);
-
/*
* The reason for kmap_high_get() is to ensure that the currently kmap'd
* page usage count does not decrease to zero while we're using its
@@ -62,7 +60,6 @@ static inline void *kmap_high_get(struct page *page)
* when CONFIG_HIGHMEM is not set.
*/
#ifdef CONFIG_HIGHMEM
-extern void kunmap(struct page *page);
extern void *kmap_atomic(struct page *page);
extern void __kunmap_atomic(void *kvaddr);
extern void *kmap_atomic_pfn(unsigned long pfn);
diff --git a/arch/arm/mm/highmem.c b/arch/arm/mm/highmem.c
index e8ba37c36590..c700b32350ee 100644
--- a/arch/arm/mm/highmem.c
+++ b/arch/arm/mm/highmem.c
@@ -31,15 +31,6 @@ static inline pte_t get_fixmap_pte(unsigned long vaddr)
return *ptep;
}
-void kunmap(struct page *page)
-{
- might_sleep();
- if (!PageHighMem(page))
- return;
- kunmap_high(page);
-}
-EXPORT_SYMBOL(kunmap);
-
void *kmap_atomic(struct page *page)
{
unsigned int idx;
diff --git a/arch/csky/include/asm/highmem.h b/arch/csky/include/asm/highmem.h
index 9d0516e38110..be11c5b67122 100644
--- a/arch/csky/include/asm/highmem.h
+++ b/arch/csky/include/asm/highmem.h
@@ -30,11 +30,8 @@ extern pte_t *pkmap_page_table;
#define PKMAP_NR(virt) ((virt-PKMAP_BASE) >> PAGE_SHIFT)
#define PKMAP_ADDR(nr) (PKMAP_BASE + ((nr) << PAGE_SHIFT))
-extern void kunmap_high(struct page *page);
-
#define ARCH_HAS_KMAP_FLUSH_TLB
extern void kmap_flush_tlb(unsigned long addr);
-extern void kunmap(struct page *page);
extern void *kmap_atomic(struct page *page);
extern void __kunmap_atomic(void *kvaddr);
extern void *kmap_atomic_pfn(unsigned long pfn);
diff --git a/arch/csky/mm/highmem.c b/arch/csky/mm/highmem.c
index 4a3c273bc8b9..e9952211264b 100644
--- a/arch/csky/mm/highmem.c
+++ b/arch/csky/mm/highmem.c
@@ -21,15 +21,6 @@ EXPORT_SYMBOL(kmap_flush_tlb);
EXPORT_SYMBOL(kmap);
-void kunmap(struct page *page)
-{
- might_sleep();
- if (!PageHighMem(page))
- return;
- kunmap_high(page);
-}
-EXPORT_SYMBOL(kunmap);
-
void *kmap_atomic(struct page *page)
{
unsigned long vaddr;
diff --git a/arch/microblaze/include/asm/highmem.h b/arch/microblaze/include/asm/highmem.h
index 8c5bfd228bd8..0c94046f2d58 100644
--- a/arch/microblaze/include/asm/highmem.h
+++ b/arch/microblaze/include/asm/highmem.h
@@ -51,18 +51,9 @@ extern pte_t *pkmap_page_table;
#define PKMAP_NR(virt) ((virt - PKMAP_BASE) >> PAGE_SHIFT)
#define PKMAP_ADDR(nr) (PKMAP_BASE + ((nr) << PAGE_SHIFT))
-extern void kunmap_high(struct page *page);
extern void *kmap_atomic_prot(struct page *page, pgprot_t prot);
extern void __kunmap_atomic(void *kvaddr);
-static inline void kunmap(struct page *page)
-{
- might_sleep();
- if (!PageHighMem(page))
- return;
- kunmap_high(page);
-}
-
static inline void *kmap_atomic(struct page *page)
{
return kmap_atomic_prot(page, kmap_prot);
diff --git a/arch/mips/include/asm/highmem.h b/arch/mips/include/asm/highmem.h
index 1f741e3ecabf..24e7e7e5cc7b 100644
--- a/arch/mips/include/asm/highmem.h
+++ b/arch/mips/include/asm/highmem.h
@@ -46,11 +46,8 @@ extern pte_t *pkmap_page_table;
#define PKMAP_NR(virt) ((virt-PKMAP_BASE) >> PAGE_SHIFT)
#define PKMAP_ADDR(nr) (PKMAP_BASE + ((nr) << PAGE_SHIFT))
-extern void kunmap_high(struct page *page);
-
#define ARCH_HAS_KMAP_FLUSH_TLB
extern void kmap_flush_tlb(unsigned long addr);
-extern void kunmap(struct page *page);
extern void *kmap_atomic(struct page *page);
extern void __kunmap_atomic(void *kvaddr);
extern void *kmap_atomic_pfn(unsigned long pfn);
diff --git a/arch/mips/mm/highmem.c b/arch/mips/mm/highmem.c
index c72058bfead6..eb8ec8493f2f 100644
--- a/arch/mips/mm/highmem.c
+++ b/arch/mips/mm/highmem.c
@@ -18,15 +18,6 @@ void kmap_flush_tlb(unsigned long addr)
}
EXPORT_SYMBOL(kmap_flush_tlb);
-void kunmap(struct page *page)
-{
- might_sleep();
- if (!PageHighMem(page))
- return;
- kunmap_high(page);
-}
-EXPORT_SYMBOL(kunmap);
-
/*
* kmap_atomic/kunmap_atomic is significantly faster than kmap/kunmap because
* no global lock is needed and because the kmap code must perform a global TLB
diff --git a/arch/nds32/include/asm/highmem.h b/arch/nds32/include/asm/highmem.h
index b13654a79069..c93c7368bb3f 100644
--- a/arch/nds32/include/asm/highmem.h
+++ b/arch/nds32/include/asm/highmem.h
@@ -44,8 +44,6 @@ extern unsigned long highstart_pfn, highend_pfn;
extern pte_t *pkmap_page_table;
-extern void kunmap_high(struct page *page);
-
extern void kmap_init(void);
/*
@@ -53,7 +51,6 @@ extern void kmap_init(void);
* when CONFIG_HIGHMEM is not set.
*/
#ifdef CONFIG_HIGHMEM
-extern void kunmap(struct page *page);
extern void *kmap_atomic(struct page *page);
extern void __kunmap_atomic(void *kvaddr);
extern void *kmap_atomic_pfn(unsigned long pfn);
diff --git a/arch/nds32/mm/highmem.c b/arch/nds32/mm/highmem.c
index d0cde53b84ae..f9348bec0ecb 100644
--- a/arch/nds32/mm/highmem.c
+++ b/arch/nds32/mm/highmem.c
@@ -10,16 +10,6 @@
#include <asm/fixmap.h>
#include <asm/tlbflush.h>
-void kunmap(struct page *page)
-{
- might_sleep();
- if (!PageHighMem(page))
- return;
- kunmap_high(page);
-}
-
-EXPORT_SYMBOL(kunmap);
-
void *kmap_atomic(struct page *page)
{
unsigned int idx;
diff --git a/arch/powerpc/include/asm/highmem.h b/arch/powerpc/include/asm/highmem.h
index f14e4feef6d5..ba3371977d49 100644
--- a/arch/powerpc/include/asm/highmem.h
+++ b/arch/powerpc/include/asm/highmem.h
@@ -59,18 +59,9 @@ extern pte_t *pkmap_page_table;
#define PKMAP_NR(virt) ((virt-PKMAP_BASE) >> PAGE_SHIFT)
#define PKMAP_ADDR(nr) (PKMAP_BASE + ((nr) << PAGE_SHIFT))
-extern void kunmap_high(struct page *page);
extern void *kmap_atomic_prot(struct page *page, pgprot_t prot);
extern void __kunmap_atomic(void *kvaddr);
-static inline void kunmap(struct page *page)
-{
- might_sleep();
- if (!PageHighMem(page))
- return;
- kunmap_high(page);
-}
-
static inline void *kmap_atomic(struct page *page)
{
return kmap_atomic_prot(page, kmap_prot);
diff --git a/arch/sparc/include/asm/highmem.h b/arch/sparc/include/asm/highmem.h
index 2ff1192047f7..4bdb79fed02c 100644
--- a/arch/sparc/include/asm/highmem.h
+++ b/arch/sparc/include/asm/highmem.h
@@ -50,16 +50,6 @@ void kmap_init(void) __init;
#define PKMAP_END (PKMAP_ADDR(LAST_PKMAP))
-void kunmap_high(struct page *page);
-
-static inline void kunmap(struct page *page)
-{
- might_sleep();
- if (!PageHighMem(page))
- return;
- kunmap_high(page);
-}
-
void *kmap_atomic(struct page *page);
void __kunmap_atomic(void *kvaddr);
diff --git a/arch/x86/include/asm/highmem.h b/arch/x86/include/asm/highmem.h
index c916a28a9738..90b96594d6c5 100644
--- a/arch/x86/include/asm/highmem.h
+++ b/arch/x86/include/asm/highmem.h
@@ -58,10 +58,6 @@ extern unsigned long highstart_pfn, highend_pfn;
#define PKMAP_NR(virt) ((virt-PKMAP_BASE) >> PAGE_SHIFT)
#define PKMAP_ADDR(nr) (PKMAP_BASE + ((nr) << PAGE_SHIFT))
-extern void kunmap_high(struct page *page);
-
-void kunmap(struct page *page);
-
void *kmap_atomic_prot(struct page *page, pgprot_t prot);
void *kmap_atomic(struct page *page);
void __kunmap_atomic(void *kvaddr);
diff --git a/arch/x86/mm/highmem_32.c b/arch/x86/mm/highmem_32.c
index 12591a81b85c..c4ebfd0ae401 100644
--- a/arch/x86/mm/highmem_32.c
+++ b/arch/x86/mm/highmem_32.c
@@ -4,15 +4,6 @@
#include <linux/swap.h> /* for totalram_pages */
#include <linux/memblock.h>
-void kunmap(struct page *page)
-{
- might_sleep();
- if (!PageHighMem(page))
- return;
- kunmap_high(page);
-}
-EXPORT_SYMBOL(kunmap);
-
/*
* kmap_atomic/kunmap_atomic is significantly faster than kmap/kunmap because
* no global lock is needed and because the kmap code must perform a global TLB
diff --git a/arch/xtensa/include/asm/highmem.h b/arch/xtensa/include/asm/highmem.h
index 2546b88ddecf..5a481f7def0b 100644
--- a/arch/xtensa/include/asm/highmem.h
+++ b/arch/xtensa/include/asm/highmem.h
@@ -63,16 +63,6 @@ static inline wait_queue_head_t *get_pkmap_wait_queue_head(unsigned int color)
extern pte_t *pkmap_page_table;
-void kunmap_high(struct page *page);
-
-static inline void kunmap(struct page *page)
-{
- might_sleep();
- if (!PageHighMem(page))
- return;
- kunmap_high(page);
-}
-
static inline void flush_cache_kmaps(void)
{
flush_cache_all();
diff --git a/include/linux/highmem.h b/include/linux/highmem.h
index fc3adc51254a..ae6e8cb81043 100644
--- a/include/linux/highmem.h
+++ b/include/linux/highmem.h
@@ -52,6 +52,15 @@ static inline void *kmap(struct page *page)
return addr;
}
+void kunmap_high(struct page *page);
+static inline void kunmap(struct page *page)
+{
+ might_sleep();
+ if (!PageHighMem(page))
+ return;
+ kunmap_high(page);
+}
+
/* declarations for linux/mm/highmem.c */
unsigned int nr_free_highpages(void);
extern atomic_long_t _totalhigh_pages;
--
2.25.1
^ permalink raw reply related
* [PATCH V1 03/10] arch/kmap: Remove redundant arch specific kmaps
From: ira.weiny @ 2020-04-30 20:38 UTC (permalink / raw)
To: linux-kernel, Andrew Morton, Christian Koenig, Huang Rui
Cc: Peter Zijlstra, Dave Hansen, dri-devel, James E.J. Bottomley,
Max Filippov, Paul Mackerras, H. Peter Anvin, sparclinux,
Ira Weiny, Thomas Gleixner, Helge Deller, x86, linux-csky,
Ingo Molnar, linux-snps-arc, linux-xtensa, Borislav Petkov,
Andy Lutomirski, Dan Williams, linux-arm-kernel, Chris Zankel,
Thomas Bogendoerfer, linux-parisc, linux-mips, linuxppc-dev,
David S. Miller
In-Reply-To: <20200430203845.582900-1-ira.weiny@intel.com>
From: Ira Weiny <ira.weiny@intel.com>
The kmap code for all the architectures is almost 100% identical.
Lift the common code to the core. Use ARCH_HAS_KMAP_FLUSH_TLB to
indicate if an arch defines kmap_flush_tlb() and call if if needed.
This also has the benefit of changing kmap() on a number of
architectures to be an inline call rather than an actual function.
Signed-off-by: Ira Weiny <ira.weiny@intel.com>
---
Changes from V0:
Define kmap_flush_tlb() and define it in csky/mips
---
arch/arc/include/asm/highmem.h | 2 --
arch/arc/mm/highmem.c | 10 ----------
arch/arm/include/asm/highmem.h | 2 --
arch/arm/mm/highmem.c | 9 ---------
arch/csky/include/asm/highmem.h | 4 ++--
arch/csky/mm/highmem.c | 14 ++++----------
arch/microblaze/include/asm/highmem.h | 9 ---------
arch/mips/include/asm/highmem.h | 4 ++--
arch/mips/mm/highmem.c | 14 +++-----------
arch/nds32/include/asm/highmem.h | 2 --
arch/nds32/mm/highmem.c | 12 ------------
arch/powerpc/include/asm/highmem.h | 9 ---------
arch/sparc/include/asm/highmem.h | 9 ---------
arch/x86/include/asm/highmem.h | 2 --
arch/x86/mm/highmem_32.c | 9 ---------
arch/xtensa/include/asm/highmem.h | 9 ---------
include/linux/highmem.h | 18 ++++++++++++++++++
17 files changed, 29 insertions(+), 109 deletions(-)
diff --git a/arch/arc/include/asm/highmem.h b/arch/arc/include/asm/highmem.h
index 042e92921c4c..96eb67c86961 100644
--- a/arch/arc/include/asm/highmem.h
+++ b/arch/arc/include/asm/highmem.h
@@ -30,8 +30,6 @@
#include <asm/cacheflush.h>
-extern void *kmap(struct page *page);
-extern void *kmap_high(struct page *page);
extern void *kmap_atomic(struct page *page);
extern void __kunmap_atomic(void *kvaddr);
extern void kunmap_high(struct page *page);
diff --git a/arch/arc/mm/highmem.c b/arch/arc/mm/highmem.c
index 39ef7b9a3aa9..4db13a6b9f3b 100644
--- a/arch/arc/mm/highmem.c
+++ b/arch/arc/mm/highmem.c
@@ -49,16 +49,6 @@
extern pte_t * pkmap_page_table;
static pte_t * fixmap_page_table;
-void *kmap(struct page *page)
-{
- might_sleep();
- if (!PageHighMem(page))
- return page_address(page);
-
- return kmap_high(page);
-}
-EXPORT_SYMBOL(kmap);
-
void *kmap_atomic(struct page *page)
{
int idx, cpu_idx;
diff --git a/arch/arm/include/asm/highmem.h b/arch/arm/include/asm/highmem.h
index eb4e4207cd3c..c917522541de 100644
--- a/arch/arm/include/asm/highmem.h
+++ b/arch/arm/include/asm/highmem.h
@@ -20,7 +20,6 @@
extern pte_t *pkmap_page_table;
-extern void *kmap_high(struct page *page);
extern void kunmap_high(struct page *page);
/*
@@ -63,7 +62,6 @@ static inline void *kmap_high_get(struct page *page)
* when CONFIG_HIGHMEM is not set.
*/
#ifdef CONFIG_HIGHMEM
-extern void *kmap(struct page *page);
extern void kunmap(struct page *page);
extern void *kmap_atomic(struct page *page);
extern void __kunmap_atomic(void *kvaddr);
diff --git a/arch/arm/mm/highmem.c b/arch/arm/mm/highmem.c
index cc6eb79ef20c..e8ba37c36590 100644
--- a/arch/arm/mm/highmem.c
+++ b/arch/arm/mm/highmem.c
@@ -31,15 +31,6 @@ static inline pte_t get_fixmap_pte(unsigned long vaddr)
return *ptep;
}
-void *kmap(struct page *page)
-{
- might_sleep();
- if (!PageHighMem(page))
- return page_address(page);
- return kmap_high(page);
-}
-EXPORT_SYMBOL(kmap);
-
void kunmap(struct page *page)
{
might_sleep();
diff --git a/arch/csky/include/asm/highmem.h b/arch/csky/include/asm/highmem.h
index a345a2f2c22e..9d0516e38110 100644
--- a/arch/csky/include/asm/highmem.h
+++ b/arch/csky/include/asm/highmem.h
@@ -30,10 +30,10 @@ extern pte_t *pkmap_page_table;
#define PKMAP_NR(virt) ((virt-PKMAP_BASE) >> PAGE_SHIFT)
#define PKMAP_ADDR(nr) (PKMAP_BASE + ((nr) << PAGE_SHIFT))
-extern void *kmap_high(struct page *page);
extern void kunmap_high(struct page *page);
-extern void *kmap(struct page *page);
+#define ARCH_HAS_KMAP_FLUSH_TLB
+extern void kmap_flush_tlb(unsigned long addr);
extern void kunmap(struct page *page);
extern void *kmap_atomic(struct page *page);
extern void __kunmap_atomic(void *kvaddr);
diff --git a/arch/csky/mm/highmem.c b/arch/csky/mm/highmem.c
index 690d678649d1..4a3c273bc8b9 100644
--- a/arch/csky/mm/highmem.c
+++ b/arch/csky/mm/highmem.c
@@ -13,18 +13,12 @@ static pte_t *kmap_pte;
unsigned long highstart_pfn, highend_pfn;
-void *kmap(struct page *page)
+void kmap_flush_tlb(unsigned long addr)
{
- void *addr;
-
- might_sleep();
- if (!PageHighMem(page))
- return page_address(page);
- addr = kmap_high(page);
- flush_tlb_one((unsigned long)addr);
-
- return addr;
+ flush_tlb_one(addr);
}
+EXPORT_SYMBOL(kmap_flush_tlb);
+
EXPORT_SYMBOL(kmap);
void kunmap(struct page *page)
diff --git a/arch/microblaze/include/asm/highmem.h b/arch/microblaze/include/asm/highmem.h
index 99ced7278b5c..8c5bfd228bd8 100644
--- a/arch/microblaze/include/asm/highmem.h
+++ b/arch/microblaze/include/asm/highmem.h
@@ -51,19 +51,10 @@ extern pte_t *pkmap_page_table;
#define PKMAP_NR(virt) ((virt - PKMAP_BASE) >> PAGE_SHIFT)
#define PKMAP_ADDR(nr) (PKMAP_BASE + ((nr) << PAGE_SHIFT))
-extern void *kmap_high(struct page *page);
extern void kunmap_high(struct page *page);
extern void *kmap_atomic_prot(struct page *page, pgprot_t prot);
extern void __kunmap_atomic(void *kvaddr);
-static inline void *kmap(struct page *page)
-{
- might_sleep();
- if (!PageHighMem(page))
- return page_address(page);
- return kmap_high(page);
-}
-
static inline void kunmap(struct page *page)
{
might_sleep();
diff --git a/arch/mips/include/asm/highmem.h b/arch/mips/include/asm/highmem.h
index 9d84aafc33d0..1f741e3ecabf 100644
--- a/arch/mips/include/asm/highmem.h
+++ b/arch/mips/include/asm/highmem.h
@@ -46,10 +46,10 @@ extern pte_t *pkmap_page_table;
#define PKMAP_NR(virt) ((virt-PKMAP_BASE) >> PAGE_SHIFT)
#define PKMAP_ADDR(nr) (PKMAP_BASE + ((nr) << PAGE_SHIFT))
-extern void * kmap_high(struct page *page);
extern void kunmap_high(struct page *page);
-extern void *kmap(struct page *page);
+#define ARCH_HAS_KMAP_FLUSH_TLB
+extern void kmap_flush_tlb(unsigned long addr);
extern void kunmap(struct page *page);
extern void *kmap_atomic(struct page *page);
extern void __kunmap_atomic(void *kvaddr);
diff --git a/arch/mips/mm/highmem.c b/arch/mips/mm/highmem.c
index edd889f6cede..c72058bfead6 100644
--- a/arch/mips/mm/highmem.c
+++ b/arch/mips/mm/highmem.c
@@ -12,19 +12,11 @@ static pte_t *kmap_pte;
unsigned long highstart_pfn, highend_pfn;
-void *kmap(struct page *page)
+void kmap_flush_tlb(unsigned long addr)
{
- void *addr;
-
- might_sleep();
- if (!PageHighMem(page))
- return page_address(page);
- addr = kmap_high(page);
- flush_tlb_one((unsigned long)addr);
-
- return addr;
+ flush_tlb_one(addr);
}
-EXPORT_SYMBOL(kmap);
+EXPORT_SYMBOL(kmap_flush_tlb);
void kunmap(struct page *page)
{
diff --git a/arch/nds32/include/asm/highmem.h b/arch/nds32/include/asm/highmem.h
index b3a82c97ded3..b13654a79069 100644
--- a/arch/nds32/include/asm/highmem.h
+++ b/arch/nds32/include/asm/highmem.h
@@ -44,7 +44,6 @@ extern unsigned long highstart_pfn, highend_pfn;
extern pte_t *pkmap_page_table;
-extern void *kmap_high(struct page *page);
extern void kunmap_high(struct page *page);
extern void kmap_init(void);
@@ -54,7 +53,6 @@ extern void kmap_init(void);
* when CONFIG_HIGHMEM is not set.
*/
#ifdef CONFIG_HIGHMEM
-extern void *kmap(struct page *page);
extern void kunmap(struct page *page);
extern void *kmap_atomic(struct page *page);
extern void __kunmap_atomic(void *kvaddr);
diff --git a/arch/nds32/mm/highmem.c b/arch/nds32/mm/highmem.c
index 4c7c28e994ea..d0cde53b84ae 100644
--- a/arch/nds32/mm/highmem.c
+++ b/arch/nds32/mm/highmem.c
@@ -10,18 +10,6 @@
#include <asm/fixmap.h>
#include <asm/tlbflush.h>
-void *kmap(struct page *page)
-{
- unsigned long vaddr;
- might_sleep();
- if (!PageHighMem(page))
- return page_address(page);
- vaddr = (unsigned long)kmap_high(page);
- return (void *)vaddr;
-}
-
-EXPORT_SYMBOL(kmap);
-
void kunmap(struct page *page)
{
might_sleep();
diff --git a/arch/powerpc/include/asm/highmem.h b/arch/powerpc/include/asm/highmem.h
index 529512f6d65a..f14e4feef6d5 100644
--- a/arch/powerpc/include/asm/highmem.h
+++ b/arch/powerpc/include/asm/highmem.h
@@ -59,19 +59,10 @@ extern pte_t *pkmap_page_table;
#define PKMAP_NR(virt) ((virt-PKMAP_BASE) >> PAGE_SHIFT)
#define PKMAP_ADDR(nr) (PKMAP_BASE + ((nr) << PAGE_SHIFT))
-extern void *kmap_high(struct page *page);
extern void kunmap_high(struct page *page);
extern void *kmap_atomic_prot(struct page *page, pgprot_t prot);
extern void __kunmap_atomic(void *kvaddr);
-static inline void *kmap(struct page *page)
-{
- might_sleep();
- if (!PageHighMem(page))
- return page_address(page);
- return kmap_high(page);
-}
-
static inline void kunmap(struct page *page)
{
might_sleep();
diff --git a/arch/sparc/include/asm/highmem.h b/arch/sparc/include/asm/highmem.h
index 7dd2d4b3f980..2ff1192047f7 100644
--- a/arch/sparc/include/asm/highmem.h
+++ b/arch/sparc/include/asm/highmem.h
@@ -50,17 +50,8 @@ void kmap_init(void) __init;
#define PKMAP_END (PKMAP_ADDR(LAST_PKMAP))
-void *kmap_high(struct page *page);
void kunmap_high(struct page *page);
-static inline void *kmap(struct page *page)
-{
- might_sleep();
- if (!PageHighMem(page))
- return page_address(page);
- return kmap_high(page);
-}
-
static inline void kunmap(struct page *page)
{
might_sleep();
diff --git a/arch/x86/include/asm/highmem.h b/arch/x86/include/asm/highmem.h
index a8059930056d..c916a28a9738 100644
--- a/arch/x86/include/asm/highmem.h
+++ b/arch/x86/include/asm/highmem.h
@@ -58,10 +58,8 @@ extern unsigned long highstart_pfn, highend_pfn;
#define PKMAP_NR(virt) ((virt-PKMAP_BASE) >> PAGE_SHIFT)
#define PKMAP_ADDR(nr) (PKMAP_BASE + ((nr) << PAGE_SHIFT))
-extern void *kmap_high(struct page *page);
extern void kunmap_high(struct page *page);
-void *kmap(struct page *page);
void kunmap(struct page *page);
void *kmap_atomic_prot(struct page *page, pgprot_t prot);
diff --git a/arch/x86/mm/highmem_32.c b/arch/x86/mm/highmem_32.c
index 8af66382672b..12591a81b85c 100644
--- a/arch/x86/mm/highmem_32.c
+++ b/arch/x86/mm/highmem_32.c
@@ -4,15 +4,6 @@
#include <linux/swap.h> /* for totalram_pages */
#include <linux/memblock.h>
-void *kmap(struct page *page)
-{
- might_sleep();
- if (!PageHighMem(page))
- return page_address(page);
- return kmap_high(page);
-}
-EXPORT_SYMBOL(kmap);
-
void kunmap(struct page *page)
{
might_sleep();
diff --git a/arch/xtensa/include/asm/highmem.h b/arch/xtensa/include/asm/highmem.h
index a9587c85be85..2546b88ddecf 100644
--- a/arch/xtensa/include/asm/highmem.h
+++ b/arch/xtensa/include/asm/highmem.h
@@ -63,17 +63,8 @@ static inline wait_queue_head_t *get_pkmap_wait_queue_head(unsigned int color)
extern pte_t *pkmap_page_table;
-void *kmap_high(struct page *page);
void kunmap_high(struct page *page);
-static inline void *kmap(struct page *page)
-{
- might_sleep();
- if (!PageHighMem(page))
- return page_address(page);
- return kmap_high(page);
-}
-
static inline void kunmap(struct page *page)
{
might_sleep();
diff --git a/include/linux/highmem.h b/include/linux/highmem.h
index ea5cdbd8c2c3..fc3adc51254a 100644
--- a/include/linux/highmem.h
+++ b/include/linux/highmem.h
@@ -34,6 +34,24 @@ static inline void invalidate_kernel_vmap_range(void *vaddr, int size)
#ifdef CONFIG_HIGHMEM
#include <asm/highmem.h>
+#ifndef ARCH_HAS_KMAP_FLUSH_TLB
+static inline void kmap_flush_tlb(unsigned long addr) { }
+#endif
+
+void *kmap_high(struct page *page);
+static inline void *kmap(struct page *page)
+{
+ void *addr;
+
+ might_sleep();
+ if (!PageHighMem(page))
+ addr = page_address(page);
+ else
+ addr = kmap_high(page);
+ kmap_flush_tlb((unsigned long)addr);
+ return addr;
+}
+
/* declarations for linux/mm/highmem.c */
unsigned int nr_free_highpages(void);
extern atomic_long_t _totalhigh_pages;
--
2.25.1
^ permalink raw reply related
* [PATCH V1 09/10] arch/kmap: Define kmap_atomic_prot() for all arch's
From: ira.weiny @ 2020-04-30 20:38 UTC (permalink / raw)
To: linux-kernel, Andrew Morton, Christian Koenig, Huang Rui
Cc: Peter Zijlstra, Dave Hansen, dri-devel, James E.J. Bottomley,
Max Filippov, Paul Mackerras, H. Peter Anvin, sparclinux,
Ira Weiny, Thomas Gleixner, Helge Deller, x86, linux-csky,
Ingo Molnar, linux-snps-arc, linux-xtensa, Borislav Petkov,
Andy Lutomirski, Dan Williams, linux-arm-kernel, Chris Zankel,
Thomas Bogendoerfer, linux-parisc, linux-mips, linuxppc-dev,
David S. Miller
In-Reply-To: <20200430203845.582900-1-ira.weiny@intel.com>
From: Ira Weiny <ira.weiny@intel.com>
To support kmap_atomic_prot(), all architectures need to support
protections passed to their kmap_atomic_high() function. Pass
protections into kmap_atomic_high() and change the name to
kmap_atomic_high_prot() to match.
Then define kmap_atomic_prot() as a core function which calls
kmap_atomic_high_prot() when needed.
Finally, redefine kmap_atomic() as a wrapper of kmap_atomic_prot() with
the default kmap_prot exported by the architectures.
Signed-off-by: Ira Weiny <ira.weiny@intel.com>
---
arch/arc/include/asm/highmem.h | 2 +-
arch/arc/mm/highmem.c | 6 +++---
arch/arm/include/asm/highmem.h | 2 +-
arch/arm/mm/highmem.c | 6 +++---
arch/csky/include/asm/highmem.h | 2 +-
arch/csky/mm/highmem.c | 6 +++---
arch/microblaze/include/asm/highmem.h | 7 +------
arch/microblaze/mm/highmem.c | 4 ++--
arch/mips/include/asm/highmem.h | 2 +-
arch/mips/mm/highmem.c | 6 +++---
arch/nds32/include/asm/highmem.h | 2 +-
arch/nds32/mm/highmem.c | 6 +++---
arch/powerpc/include/asm/highmem.h | 8 +-------
arch/powerpc/mm/highmem.c | 4 ++--
arch/sparc/include/asm/highmem.h | 2 +-
arch/sparc/mm/highmem.c | 6 +++---
arch/x86/include/asm/highmem.h | 6 +-----
arch/x86/mm/highmem_32.c | 4 ++--
arch/xtensa/include/asm/highmem.h | 2 +-
arch/xtensa/mm/highmem.c | 6 +++---
include/linux/highmem.h | 5 +++--
21 files changed, 40 insertions(+), 54 deletions(-)
diff --git a/arch/arc/include/asm/highmem.h b/arch/arc/include/asm/highmem.h
index e16531495620..09f86bde6809 100644
--- a/arch/arc/include/asm/highmem.h
+++ b/arch/arc/include/asm/highmem.h
@@ -30,7 +30,7 @@
#include <asm/cacheflush.h>
-extern void *kmap_atomic_high(struct page *page);
+extern void *kmap_atomic_high_prot(struct page *page, pgprot_t prot);
extern void kunmap_atomic_high(void *kvaddr);
extern void kmap_init(void);
diff --git a/arch/arc/mm/highmem.c b/arch/arc/mm/highmem.c
index 5d3eab4ac0b0..479b0d72d3cf 100644
--- a/arch/arc/mm/highmem.c
+++ b/arch/arc/mm/highmem.c
@@ -49,7 +49,7 @@
extern pte_t * pkmap_page_table;
static pte_t * fixmap_page_table;
-void *kmap_atomic_high(struct page *page)
+void *kmap_atomic_high_prot(struct page *page, pgprot_t prot)
{
int idx, cpu_idx;
unsigned long vaddr;
@@ -59,11 +59,11 @@ void *kmap_atomic_high(struct page *page)
vaddr = FIXMAP_ADDR(idx);
set_pte_at(&init_mm, vaddr, fixmap_page_table + idx,
- mk_pte(page, kmap_prot));
+ mk_pte(page, prot));
return (void *)vaddr;
}
-EXPORT_SYMBOL(kmap_atomic_high);
+EXPORT_SYMBOL(kmap_atomic_high_prot);
void kunmap_atomic_high(void *kv)
{
diff --git a/arch/arm/include/asm/highmem.h b/arch/arm/include/asm/highmem.h
index a9d5e9bce1cc..e35f2f73f6aa 100644
--- a/arch/arm/include/asm/highmem.h
+++ b/arch/arm/include/asm/highmem.h
@@ -60,7 +60,7 @@ static inline void *kmap_high_get(struct page *page)
* when CONFIG_HIGHMEM is not set.
*/
#ifdef CONFIG_HIGHMEM
-extern void *kmap_atomic_high(struct page *page);
+extern void *kmap_atomic_high_prot(struct page *page, pgprot_t prot);
extern void kunmap_atomic_high(void *kvaddr);
extern void *kmap_atomic_pfn(unsigned long pfn);
#endif
diff --git a/arch/arm/mm/highmem.c b/arch/arm/mm/highmem.c
index ac8394655a6e..e013f6b81328 100644
--- a/arch/arm/mm/highmem.c
+++ b/arch/arm/mm/highmem.c
@@ -31,7 +31,7 @@ static inline pte_t get_fixmap_pte(unsigned long vaddr)
return *ptep;
}
-void *kmap_atomic_high(struct page *page)
+void *kmap_atomic_high_prot(struct page *page, pgprot_t prot)
{
unsigned int idx;
unsigned long vaddr;
@@ -67,11 +67,11 @@ void *kmap_atomic_high(struct page *page)
* in place, so the contained TLB flush ensures the TLB is updated
* with the new mapping.
*/
- set_fixmap_pte(idx, mk_pte(page, kmap_prot));
+ set_fixmap_pte(idx, mk_pte(page, prot));
return (void *)vaddr;
}
-EXPORT_SYMBOL(kmap_atomic_high);
+EXPORT_SYMBOL(kmap_atomic_high_prot);
void kunmap_atomic_high(void *kvaddr)
{
diff --git a/arch/csky/include/asm/highmem.h b/arch/csky/include/asm/highmem.h
index 5bbbe59e60a9..59854c7ccf78 100644
--- a/arch/csky/include/asm/highmem.h
+++ b/arch/csky/include/asm/highmem.h
@@ -32,7 +32,7 @@ extern pte_t *pkmap_page_table;
#define ARCH_HAS_KMAP_FLUSH_TLB
extern void kmap_flush_tlb(unsigned long addr);
-extern void *kmap_atomic_high(struct page *page);
+extern void *kmap_atomic_high_prot(struct page *page, pgprot_t prot);
extern void kunmap_atomic_high(void *kvaddr);
extern void *kmap_atomic_pfn(unsigned long pfn);
extern struct page *kmap_atomic_to_page(void *ptr);
diff --git a/arch/csky/mm/highmem.c b/arch/csky/mm/highmem.c
index f4311669b5bb..3ae5c8cd7619 100644
--- a/arch/csky/mm/highmem.c
+++ b/arch/csky/mm/highmem.c
@@ -21,7 +21,7 @@ EXPORT_SYMBOL(kmap_flush_tlb);
EXPORT_SYMBOL(kmap);
-void *kmap_atomic_high(struct page *page)
+void *kmap_atomic_high_prot(struct page *page, pgprot_t prot)
{
unsigned long vaddr;
int idx, type;
@@ -32,12 +32,12 @@ void *kmap_atomic_high(struct page *page)
#ifdef CONFIG_DEBUG_HIGHMEM
BUG_ON(!pte_none(*(kmap_pte - idx)));
#endif
- set_pte(kmap_pte-idx, mk_pte(page, kmap_prot));
+ set_pte(kmap_pte-idx, mk_pte(page, prot));
flush_tlb_one((unsigned long)vaddr);
return (void *)vaddr;
}
-EXPORT_SYMBOL(kmap_atomic_high);
+EXPORT_SYMBOL(kmap_atomic_high_prot);
void kunmap_atomic_high(void *kvaddr)
{
diff --git a/arch/microblaze/include/asm/highmem.h b/arch/microblaze/include/asm/highmem.h
index 66521fdc3a47..eb0a2cb883bd 100644
--- a/arch/microblaze/include/asm/highmem.h
+++ b/arch/microblaze/include/asm/highmem.h
@@ -51,14 +51,9 @@ extern pte_t *pkmap_page_table;
#define PKMAP_NR(virt) ((virt - PKMAP_BASE) >> PAGE_SHIFT)
#define PKMAP_ADDR(nr) (PKMAP_BASE + ((nr) << PAGE_SHIFT))
-extern void *kmap_atomic_prot(struct page *page, pgprot_t prot);
+extern void *kmap_atomic_high_prot(struct page *page, pgprot_t prot);
extern void kunmap_atomic_high(void *kvaddr);
-static inline void *kmap_atomic_high(struct page *page)
-{
- return kmap_atomic_prot(page, kmap_prot);
-}
-
#define flush_cache_kmaps() { flush_icache(); flush_dcache(); }
#endif /* __KERNEL__ */
diff --git a/arch/microblaze/mm/highmem.c b/arch/microblaze/mm/highmem.c
index 1026aeffe11a..ee8a422b2b76 100644
--- a/arch/microblaze/mm/highmem.c
+++ b/arch/microblaze/mm/highmem.c
@@ -32,7 +32,7 @@
*/
#include <asm/tlbflush.h>
-void *kmap_atomic_prot(struct page *page, pgprot_t prot)
+void *kmap_atomic_high_prot(struct page *page, pgprot_t prot)
{
unsigned long vaddr;
@@ -49,7 +49,7 @@ void *kmap_atomic_prot(struct page *page, pgprot_t prot)
return (void *) vaddr;
}
-EXPORT_SYMBOL(kmap_atomic_prot);
+EXPORT_SYMBOL(kmap_atomic_high_prot);
void kunmap_atomic_high(void *kvaddr)
{
diff --git a/arch/mips/include/asm/highmem.h b/arch/mips/include/asm/highmem.h
index d9f774bd4938..c9f46b450a68 100644
--- a/arch/mips/include/asm/highmem.h
+++ b/arch/mips/include/asm/highmem.h
@@ -48,7 +48,7 @@ extern pte_t *pkmap_page_table;
#define ARCH_HAS_KMAP_FLUSH_TLB
extern void kmap_flush_tlb(unsigned long addr);
-extern void *kmap_atomic_high(struct page *page);
+extern void *kmap_atomic_high_prot(struct page *page, pgprot_t prot);
extern void kunmap_atomic_high(void *kvaddr);
extern void *kmap_atomic_pfn(unsigned long pfn);
diff --git a/arch/mips/mm/highmem.c b/arch/mips/mm/highmem.c
index 87023bd1a33c..37e244cdb14e 100644
--- a/arch/mips/mm/highmem.c
+++ b/arch/mips/mm/highmem.c
@@ -18,7 +18,7 @@ void kmap_flush_tlb(unsigned long addr)
}
EXPORT_SYMBOL(kmap_flush_tlb);
-void *kmap_atomic_high(struct page *page)
+void *kmap_atomic_high_prot(struct page *page, pgprot_t prot)
{
unsigned long vaddr;
int idx, type;
@@ -29,12 +29,12 @@ void *kmap_atomic_high(struct page *page)
#ifdef CONFIG_DEBUG_HIGHMEM
BUG_ON(!pte_none(*(kmap_pte - idx)));
#endif
- set_pte(kmap_pte-idx, mk_pte(page, kmap_prot));
+ set_pte(kmap_pte-idx, mk_pte(page, prot));
local_flush_tlb_one((unsigned long)vaddr);
return (void*) vaddr;
}
-EXPORT_SYMBOL(kmap_atomic_high);
+EXPORT_SYMBOL(kmap_atomic_high_prot);
void kunmap_atomic_high(void *kvaddr)
{
diff --git a/arch/nds32/include/asm/highmem.h b/arch/nds32/include/asm/highmem.h
index 97648b678108..1f9fc74d112d 100644
--- a/arch/nds32/include/asm/highmem.h
+++ b/arch/nds32/include/asm/highmem.h
@@ -51,7 +51,7 @@ extern void kmap_init(void);
* when CONFIG_HIGHMEM is not set.
*/
#ifdef CONFIG_HIGHMEM
-extern void *kmap_atomic_high(struct page *page);
+extern void *kmap_atomic_high_prot(struct page *page, pgprot_t prot);
extern void kunmap_atomic_high(void *kvaddr);
extern void *kmap_atomic_pfn(unsigned long pfn);
extern struct page *kmap_atomic_to_page(void *ptr);
diff --git a/arch/nds32/mm/highmem.c b/arch/nds32/mm/highmem.c
index 809f8c830f06..63ded527c1e8 100644
--- a/arch/nds32/mm/highmem.c
+++ b/arch/nds32/mm/highmem.c
@@ -10,7 +10,7 @@
#include <asm/fixmap.h>
#include <asm/tlbflush.h>
-void *kmap_atomic_high(struct page *page)
+void *kmap_atomic_high_prot(struct page *page, pgprot_t prot)
{
unsigned int idx;
unsigned long vaddr, pte;
@@ -21,7 +21,7 @@ void *kmap_atomic_high(struct page *page)
idx = type + KM_TYPE_NR * smp_processor_id();
vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
- pte = (page_to_pfn(page) << PAGE_SHIFT) | (kmap_prot);
+ pte = (page_to_pfn(page) << PAGE_SHIFT) | prot;
ptep = pte_offset_kernel(pmd_off_k(vaddr), vaddr);
set_pte(ptep, pte);
@@ -32,7 +32,7 @@ void *kmap_atomic_high(struct page *page)
return (void *)vaddr;
}
-EXPORT_SYMBOL(kmap_atomic_high);
+EXPORT_SYMBOL(kmap_atomic_high_prot);
void kunmap_atomic_high(void *kvaddr)
{
diff --git a/arch/powerpc/include/asm/highmem.h b/arch/powerpc/include/asm/highmem.h
index d264aebcaa9b..edd01bbe5a44 100644
--- a/arch/powerpc/include/asm/highmem.h
+++ b/arch/powerpc/include/asm/highmem.h
@@ -59,15 +59,9 @@ extern pte_t *pkmap_page_table;
#define PKMAP_NR(virt) ((virt-PKMAP_BASE) >> PAGE_SHIFT)
#define PKMAP_ADDR(nr) (PKMAP_BASE + ((nr) << PAGE_SHIFT))
-extern void *kmap_atomic_prot(struct page *page, pgprot_t prot);
+extern void *kmap_atomic_high_prot(struct page *page, pgprot_t prot);
extern void kunmap_atomic_high(void *kvaddr);
-static inline void *kmap_atomic_high(struct page *page)
-{
- return kmap_atomic_prot(page, kmap_prot);
-}
-
-
#define flush_cache_kmaps() flush_cache_all()
#endif /* __KERNEL__ */
diff --git a/arch/powerpc/mm/highmem.c b/arch/powerpc/mm/highmem.c
index 162958321e28..35071c2913f1 100644
--- a/arch/powerpc/mm/highmem.c
+++ b/arch/powerpc/mm/highmem.c
@@ -24,7 +24,7 @@
#include <linux/highmem.h>
#include <linux/module.h>
-void *kmap_atomic_prot(struct page *page, pgprot_t prot)
+void *kmap_atomic_high_prot(struct page *page, pgprot_t prot)
{
unsigned long vaddr;
int idx, type;
@@ -38,7 +38,7 @@ void *kmap_atomic_prot(struct page *page, pgprot_t prot)
return (void*) vaddr;
}
-EXPORT_SYMBOL(kmap_atomic_prot);
+EXPORT_SYMBOL(kmap_atomic_high_prot);
void kunmap_atomic_high(void *kvaddr)
{
diff --git a/arch/sparc/include/asm/highmem.h b/arch/sparc/include/asm/highmem.h
index 94dd6e4c5fa4..d5c5700672de 100644
--- a/arch/sparc/include/asm/highmem.h
+++ b/arch/sparc/include/asm/highmem.h
@@ -50,7 +50,7 @@ void kmap_init(void) __init;
#define PKMAP_END (PKMAP_ADDR(LAST_PKMAP))
-void *kmap_atomic_high(struct page *page);
+void *kmap_atomic_high_prot(struct page *page, pgprot_t prot);
void kunmap_atomic_high(void *kvaddr);
#define flush_cache_kmaps() flush_cache_all()
diff --git a/arch/sparc/mm/highmem.c b/arch/sparc/mm/highmem.c
index 9f06d75e88e1..414f578d1e57 100644
--- a/arch/sparc/mm/highmem.c
+++ b/arch/sparc/mm/highmem.c
@@ -54,7 +54,7 @@ void __init kmap_init(void)
kmap_prot = __pgprot(SRMMU_ET_PTE | SRMMU_PRIV | SRMMU_CACHE);
}
-void *kmap_atomic_high(struct page *page)
+void *kmap_atomic_high_prot(struct page *page, pgprot_t prot)
{
unsigned long vaddr;
long idx, type;
@@ -73,7 +73,7 @@ void *kmap_atomic_high(struct page *page)
#ifdef CONFIG_DEBUG_HIGHMEM
BUG_ON(!pte_none(*(kmap_pte-idx)));
#endif
- set_pte(kmap_pte-idx, mk_pte(page, kmap_prot));
+ set_pte(kmap_pte-idx, mk_pte(page, prot));
/* XXX Fix - Anton */
#if 0
__flush_tlb_one(vaddr);
@@ -83,7 +83,7 @@ void *kmap_atomic_high(struct page *page)
return (void*) vaddr;
}
-EXPORT_SYMBOL(kmap_atomic_high);
+EXPORT_SYMBOL(kmap_atomic_high_prot);
void kunmap_atomic_high(void *kvaddr)
{
diff --git a/arch/x86/include/asm/highmem.h b/arch/x86/include/asm/highmem.h
index ff87aba96eee..009b8e22e906 100644
--- a/arch/x86/include/asm/highmem.h
+++ b/arch/x86/include/asm/highmem.h
@@ -58,11 +58,7 @@ extern unsigned long highstart_pfn, highend_pfn;
#define PKMAP_NR(virt) ((virt-PKMAP_BASE) >> PAGE_SHIFT)
#define PKMAP_ADDR(nr) (PKMAP_BASE + ((nr) << PAGE_SHIFT))
-extern void *kmap_atomic_prot(struct page *page, pgprot_t prot);
-void *kmap_atomic_high(struct page *page)
-{
- return kmap_atomic_prot(page, kmap_prot);
-}
+void *kmap_atomic_high_prot(struct page *page, pgprot_t prot);
void kunmap_atomic_high(void *kvaddr);
void *kmap_atomic_pfn(unsigned long pfn);
void *kmap_atomic_prot_pfn(unsigned long pfn, pgprot_t prot);
diff --git a/arch/x86/mm/highmem_32.c b/arch/x86/mm/highmem_32.c
index af07a6842743..075fe51317b0 100644
--- a/arch/x86/mm/highmem_32.c
+++ b/arch/x86/mm/highmem_32.c
@@ -4,7 +4,7 @@
#include <linux/swap.h> /* for totalram_pages */
#include <linux/memblock.h>
-void *kmap_atomic_prot(struct page *page, pgprot_t prot)
+void *kmap_atomic_high_prot(struct page *page, pgprot_t prot)
{
unsigned long vaddr;
int idx, type;
@@ -18,7 +18,7 @@ void *kmap_atomic_prot(struct page *page, pgprot_t prot)
return (void *)vaddr;
}
-EXPORT_SYMBOL(kmap_atomic_prot);
+EXPORT_SYMBOL(kmap_atomic_high_prot);
/*
* This is the same as kmap_atomic() but can map memory that doesn't
diff --git a/arch/xtensa/include/asm/highmem.h b/arch/xtensa/include/asm/highmem.h
index a60a02dc68f6..7152aeb1e3a4 100644
--- a/arch/xtensa/include/asm/highmem.h
+++ b/arch/xtensa/include/asm/highmem.h
@@ -68,7 +68,7 @@ static inline void flush_cache_kmaps(void)
flush_cache_all();
}
-void *kmap_atomic_high(struct page *page);
+void *kmap_atomic_high_prot(struct page *page, pgprot_t prot);
void kunmap_atomic_high(void *kvaddr);
void kmap_init(void);
diff --git a/arch/xtensa/mm/highmem.c b/arch/xtensa/mm/highmem.c
index 8c58c4c37033..fe56644d7b23 100644
--- a/arch/xtensa/mm/highmem.c
+++ b/arch/xtensa/mm/highmem.c
@@ -37,7 +37,7 @@ static inline enum fixed_addresses kmap_idx(int type, unsigned long color)
color;
}
-void *kmap_atomic_high(struct page *page)
+void *kmap_atomic_high_prot(struct page *page, pgprot_t prot)
{
enum fixed_addresses idx;
unsigned long vaddr;
@@ -48,11 +48,11 @@ void *kmap_atomic_high(struct page *page)
#ifdef CONFIG_DEBUG_HIGHMEM
BUG_ON(!pte_none(*(kmap_pte + idx)));
#endif
- set_pte(kmap_pte + idx, mk_pte(page, kmap_prot));
+ set_pte(kmap_pte + idx, mk_pte(page, prot));
return (void *)vaddr;
}
-EXPORT_SYMBOL(kmap_atomic_high);
+EXPORT_SYMBOL(kmap_atomic_high_prot);
void kunmap_atomic_high(void *kvaddr)
{
diff --git a/include/linux/highmem.h b/include/linux/highmem.h
index 601df07607a4..b10e8a39ae60 100644
--- a/include/linux/highmem.h
+++ b/include/linux/highmem.h
@@ -74,14 +74,15 @@ static inline void kunmap(struct page *page)
* be used in IRQ contexts, so in some (very limited) cases we need
* it.
*/
-static inline void *kmap_atomic(struct page *page)
+static inline void *kmap_atomic_prot(struct page *page, pgprot_t prot)
{
preempt_disable();
pagefault_disable();
if (!PageHighMem(page))
return page_address(page);
- return kmap_atomic_high(page);
+ return kmap_atomic_high_prot(page, prot);
}
+#define kmap_atomic(page) kmap_atomic_prot(page, kmap_prot)
/* declarations for linux/mm/highmem.c */
unsigned int nr_free_highpages(void);
--
2.25.1
^ permalink raw reply related
* [PATCH V1 10/10] drm: Remove drm specific kmap_atomic code
From: ira.weiny @ 2020-04-30 20:38 UTC (permalink / raw)
To: linux-kernel, Andrew Morton, Christian Koenig, Huang Rui
Cc: Peter Zijlstra, Dave Hansen, dri-devel, James E.J. Bottomley,
Max Filippov, Paul Mackerras, H. Peter Anvin, sparclinux,
Ira Weiny, Thomas Gleixner, Helge Deller, x86, linux-csky,
Ingo Molnar, linux-snps-arc, linux-xtensa, Borislav Petkov,
Andy Lutomirski, Dan Williams, linux-arm-kernel, Chris Zankel,
Thomas Bogendoerfer, linux-parisc, linux-mips, linuxppc-dev,
David S. Miller
In-Reply-To: <20200430203845.582900-1-ira.weiny@intel.com>
From: Ira Weiny <ira.weiny@intel.com>
kmap_atomic_prot() is now exported by all architectures. Use this
function rather than open coding a driver specific kmap_atomic.
Signed-off-by: Ira Weiny <ira.weiny@intel.com>
---
drivers/gpu/drm/ttm/ttm_bo_util.c | 56 ++--------------------------
drivers/gpu/drm/vmwgfx/vmwgfx_blit.c | 16 ++++----
include/drm/ttm/ttm_bo_api.h | 4 --
3 files changed, 12 insertions(+), 64 deletions(-)
diff --git a/drivers/gpu/drm/ttm/ttm_bo_util.c b/drivers/gpu/drm/ttm/ttm_bo_util.c
index 52d2b71f1588..f09b096ba4fd 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_util.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_util.c
@@ -257,54 +257,6 @@ static int ttm_copy_io_page(void *dst, void *src, unsigned long page)
return 0;
}
-#ifdef CONFIG_X86
-#define __ttm_kmap_atomic_prot(__page, __prot) kmap_atomic_prot(__page, __prot)
-#define __ttm_kunmap_atomic(__addr) kunmap_atomic(__addr)
-#else
-#define __ttm_kmap_atomic_prot(__page, __prot) vmap(&__page, 1, 0, __prot)
-#define __ttm_kunmap_atomic(__addr) vunmap(__addr)
-#endif
-
-
-/**
- * ttm_kmap_atomic_prot - Efficient kernel map of a single page with
- * specified page protection.
- *
- * @page: The page to map.
- * @prot: The page protection.
- *
- * This function maps a TTM page using the kmap_atomic api if available,
- * otherwise falls back to vmap. The user must make sure that the
- * specified page does not have an aliased mapping with a different caching
- * policy unless the architecture explicitly allows it. Also mapping and
- * unmapping using this api must be correctly nested. Unmapping should
- * occur in the reverse order of mapping.
- */
-void *ttm_kmap_atomic_prot(struct page *page, pgprot_t prot)
-{
- if (pgprot_val(prot) == pgprot_val(PAGE_KERNEL))
- return kmap_atomic(page);
- else
- return __ttm_kmap_atomic_prot(page, prot);
-}
-EXPORT_SYMBOL(ttm_kmap_atomic_prot);
-
-/**
- * ttm_kunmap_atomic_prot - Unmap a page that was mapped using
- * ttm_kmap_atomic_prot.
- *
- * @addr: The virtual address from the map.
- * @prot: The page protection.
- */
-void ttm_kunmap_atomic_prot(void *addr, pgprot_t prot)
-{
- if (pgprot_val(prot) == pgprot_val(PAGE_KERNEL))
- kunmap_atomic(addr);
- else
- __ttm_kunmap_atomic(addr);
-}
-EXPORT_SYMBOL(ttm_kunmap_atomic_prot);
-
static int ttm_copy_io_ttm_page(struct ttm_tt *ttm, void *src,
unsigned long page,
pgprot_t prot)
@@ -316,13 +268,13 @@ static int ttm_copy_io_ttm_page(struct ttm_tt *ttm, void *src,
return -ENOMEM;
src = (void *)((unsigned long)src + (page << PAGE_SHIFT));
- dst = ttm_kmap_atomic_prot(d, prot);
+ dst = kmap_atomic_prot(d, prot);
if (!dst)
return -ENOMEM;
memcpy_fromio(dst, src, PAGE_SIZE);
- ttm_kunmap_atomic_prot(dst, prot);
+ kunmap_atomic(dst);
return 0;
}
@@ -338,13 +290,13 @@ static int ttm_copy_ttm_io_page(struct ttm_tt *ttm, void *dst,
return -ENOMEM;
dst = (void *)((unsigned long)dst + (page << PAGE_SHIFT));
- src = ttm_kmap_atomic_prot(s, prot);
+ src = kmap_atomic_prot(s, prot);
if (!src)
return -ENOMEM;
memcpy_toio(dst, src, PAGE_SIZE);
- ttm_kunmap_atomic_prot(src, prot);
+ kunmap_atomic(src);
return 0;
}
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_blit.c b/drivers/gpu/drm/vmwgfx/vmwgfx_blit.c
index bb46ca0c458f..94d456a1d1a9 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_blit.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_blit.c
@@ -374,12 +374,12 @@ static int vmw_bo_cpu_blit_line(struct vmw_bo_blit_line_data *d,
copy_size = min_t(u32, copy_size, PAGE_SIZE - src_page_offset);
if (unmap_src) {
- ttm_kunmap_atomic_prot(d->src_addr, d->src_prot);
+ kunmap_atomic(d->src_addr);
d->src_addr = NULL;
}
if (unmap_dst) {
- ttm_kunmap_atomic_prot(d->dst_addr, d->dst_prot);
+ kunmap_atomic(d->dst_addr);
d->dst_addr = NULL;
}
@@ -388,8 +388,8 @@ static int vmw_bo_cpu_blit_line(struct vmw_bo_blit_line_data *d,
return -EINVAL;
d->dst_addr =
- ttm_kmap_atomic_prot(d->dst_pages[dst_page],
- d->dst_prot);
+ kmap_atomic_prot(d->dst_pages[dst_page],
+ d->dst_prot);
if (!d->dst_addr)
return -ENOMEM;
@@ -401,8 +401,8 @@ static int vmw_bo_cpu_blit_line(struct vmw_bo_blit_line_data *d,
return -EINVAL;
d->src_addr =
- ttm_kmap_atomic_prot(d->src_pages[src_page],
- d->src_prot);
+ kmap_atomic_prot(d->src_pages[src_page],
+ d->src_prot);
if (!d->src_addr)
return -ENOMEM;
@@ -499,9 +499,9 @@ int vmw_bo_cpu_blit(struct ttm_buffer_object *dst,
}
out:
if (d.src_addr)
- ttm_kunmap_atomic_prot(d.src_addr, d.src_prot);
+ kunmap_atomic(d.src_addr);
if (d.dst_addr)
- ttm_kunmap_atomic_prot(d.dst_addr, d.dst_prot);
+ kunmap_atomic(d.dst_addr);
return ret;
}
diff --git a/include/drm/ttm/ttm_bo_api.h b/include/drm/ttm/ttm_bo_api.h
index 0a9d042e075a..de1ccdcd5703 100644
--- a/include/drm/ttm/ttm_bo_api.h
+++ b/include/drm/ttm/ttm_bo_api.h
@@ -668,10 +668,6 @@ int ttm_bo_mmap_obj(struct vm_area_struct *vma, struct ttm_buffer_object *bo);
int ttm_bo_mmap(struct file *filp, struct vm_area_struct *vma,
struct ttm_bo_device *bdev);
-void *ttm_kmap_atomic_prot(struct page *page, pgprot_t prot);
-
-void ttm_kunmap_atomic_prot(void *addr, pgprot_t prot);
-
/**
* ttm_bo_io
*
--
2.25.1
^ permalink raw reply related
* [PATCH V1 07/10] arch/kmap: Ensure kmap_prot visibility
From: ira.weiny @ 2020-04-30 20:38 UTC (permalink / raw)
To: linux-kernel, Andrew Morton, Christian Koenig, Huang Rui
Cc: Peter Zijlstra, Dave Hansen, dri-devel, James E.J. Bottomley,
Max Filippov, Paul Mackerras, H. Peter Anvin, sparclinux,
Ira Weiny, Thomas Gleixner, Helge Deller, x86, linux-csky,
Ingo Molnar, linux-snps-arc, linux-xtensa, Borislav Petkov,
Andy Lutomirski, Dan Williams, linux-arm-kernel, Chris Zankel,
Thomas Bogendoerfer, linux-parisc, linux-mips, linuxppc-dev,
David S. Miller
In-Reply-To: <20200430203845.582900-1-ira.weiny@intel.com>
From: Ira Weiny <ira.weiny@intel.com>
We want to support kmap_atomic_prot() on all architectures and it makes
sense to define kmap_atomic() to use the default kmap_prot.
So we ensure all arch's have a globally available kmap_prot either as a
define or exported symbol.
Signed-off-by: Ira Weiny <ira.weiny@intel.com>
---
arch/microblaze/include/asm/highmem.h | 2 +-
arch/microblaze/mm/init.c | 3 ---
arch/powerpc/include/asm/highmem.h | 2 +-
arch/powerpc/mm/mem.c | 3 ---
arch/sparc/mm/highmem.c | 1 +
5 files changed, 3 insertions(+), 8 deletions(-)
diff --git a/arch/microblaze/include/asm/highmem.h b/arch/microblaze/include/asm/highmem.h
index 5fc56b0107be..66521fdc3a47 100644
--- a/arch/microblaze/include/asm/highmem.h
+++ b/arch/microblaze/include/asm/highmem.h
@@ -25,8 +25,8 @@
#include <linux/uaccess.h>
#include <asm/fixmap.h>
+#define kmap_prot PAGE_KERNEL
extern pte_t *kmap_pte;
-extern pgprot_t kmap_prot;
extern pte_t *pkmap_page_table;
/*
diff --git a/arch/microblaze/mm/init.c b/arch/microblaze/mm/init.c
index 1ffbfa96b9b8..a467686c13af 100644
--- a/arch/microblaze/mm/init.c
+++ b/arch/microblaze/mm/init.c
@@ -49,8 +49,6 @@ unsigned long lowmem_size;
#ifdef CONFIG_HIGHMEM
pte_t *kmap_pte;
EXPORT_SYMBOL(kmap_pte);
-pgprot_t kmap_prot;
-EXPORT_SYMBOL(kmap_prot);
static inline pte_t *virt_to_kpte(unsigned long vaddr)
{
@@ -68,7 +66,6 @@ static void __init highmem_init(void)
pkmap_page_table = virt_to_kpte(PKMAP_BASE);
kmap_pte = virt_to_kpte(__fix_to_virt(FIX_KMAP_BEGIN));
- kmap_prot = PAGE_KERNEL;
}
static void highmem_setup(void)
diff --git a/arch/powerpc/include/asm/highmem.h b/arch/powerpc/include/asm/highmem.h
index 1845fbd7ce61..d264aebcaa9b 100644
--- a/arch/powerpc/include/asm/highmem.h
+++ b/arch/powerpc/include/asm/highmem.h
@@ -29,8 +29,8 @@
#include <asm/page.h>
#include <asm/fixmap.h>
+#define kmap_prot PAGE_KERNEL
extern pte_t *kmap_pte;
-extern pgprot_t kmap_prot;
extern pte_t *pkmap_page_table;
/*
diff --git a/arch/powerpc/mm/mem.c b/arch/powerpc/mm/mem.c
index 041ed7cfd341..3f642b058731 100644
--- a/arch/powerpc/mm/mem.c
+++ b/arch/powerpc/mm/mem.c
@@ -64,8 +64,6 @@ bool init_mem_is_free;
#ifdef CONFIG_HIGHMEM
pte_t *kmap_pte;
EXPORT_SYMBOL(kmap_pte);
-pgprot_t kmap_prot;
-EXPORT_SYMBOL(kmap_prot);
#endif
pgprot_t phys_mem_access_prot(struct file *file, unsigned long pfn,
@@ -245,7 +243,6 @@ void __init paging_init(void)
pkmap_page_table = virt_to_kpte(PKMAP_BASE);
kmap_pte = virt_to_kpte(__fix_to_virt(FIX_KMAP_BEGIN));
- kmap_prot = PAGE_KERNEL;
#endif /* CONFIG_HIGHMEM */
printk(KERN_DEBUG "Top of RAM: 0x%llx, Total RAM: 0x%llx\n",
diff --git a/arch/sparc/mm/highmem.c b/arch/sparc/mm/highmem.c
index 469786bc430f..9f06d75e88e1 100644
--- a/arch/sparc/mm/highmem.c
+++ b/arch/sparc/mm/highmem.c
@@ -33,6 +33,7 @@
#include <asm/vaddrs.h>
pgprot_t kmap_prot;
+EXPORT_SYMBOL(kmap_prot);
static pte_t *kmap_pte;
--
2.25.1
^ permalink raw reply related
* [PATCH V1 08/10] arch/kmap: Don't hard code kmap_prot values
From: ira.weiny @ 2020-04-30 20:38 UTC (permalink / raw)
To: linux-kernel, Andrew Morton, Christian Koenig, Huang Rui
Cc: Peter Zijlstra, Dave Hansen, dri-devel, James E.J. Bottomley,
Max Filippov, Paul Mackerras, H. Peter Anvin, sparclinux,
Ira Weiny, Thomas Gleixner, Helge Deller, x86, linux-csky,
Ingo Molnar, linux-snps-arc, linux-xtensa, Borislav Petkov,
Andy Lutomirski, Dan Williams, linux-arm-kernel, Chris Zankel,
Thomas Bogendoerfer, linux-parisc, linux-mips, linuxppc-dev,
David S. Miller
In-Reply-To: <20200430203845.582900-1-ira.weiny@intel.com>
From: Ira Weiny <ira.weiny@intel.com>
To support kmap_atomic_prot() on all architectures each arch must
support protections passed in to them.
Change csky, mips, nds32 and xtensa to use their global kmap_prot value
rather than a hard coded value which was equal.
Signed-off-by: Ira Weiny <ira.weiny@intel.com>
---
arch/csky/mm/highmem.c | 2 +-
arch/mips/mm/highmem.c | 2 +-
arch/nds32/mm/highmem.c | 2 +-
arch/xtensa/mm/highmem.c | 2 +-
4 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/arch/csky/mm/highmem.c b/arch/csky/mm/highmem.c
index 0aafbbbe651c..f4311669b5bb 100644
--- a/arch/csky/mm/highmem.c
+++ b/arch/csky/mm/highmem.c
@@ -32,7 +32,7 @@ void *kmap_atomic_high(struct page *page)
#ifdef CONFIG_DEBUG_HIGHMEM
BUG_ON(!pte_none(*(kmap_pte - idx)));
#endif
- set_pte(kmap_pte-idx, mk_pte(page, PAGE_KERNEL));
+ set_pte(kmap_pte-idx, mk_pte(page, kmap_prot));
flush_tlb_one((unsigned long)vaddr);
return (void *)vaddr;
diff --git a/arch/mips/mm/highmem.c b/arch/mips/mm/highmem.c
index 155fbb107b35..87023bd1a33c 100644
--- a/arch/mips/mm/highmem.c
+++ b/arch/mips/mm/highmem.c
@@ -29,7 +29,7 @@ void *kmap_atomic_high(struct page *page)
#ifdef CONFIG_DEBUG_HIGHMEM
BUG_ON(!pte_none(*(kmap_pte - idx)));
#endif
- set_pte(kmap_pte-idx, mk_pte(page, PAGE_KERNEL));
+ set_pte(kmap_pte-idx, mk_pte(page, kmap_prot));
local_flush_tlb_one((unsigned long)vaddr);
return (void*) vaddr;
diff --git a/arch/nds32/mm/highmem.c b/arch/nds32/mm/highmem.c
index f6e6915c0d31..809f8c830f06 100644
--- a/arch/nds32/mm/highmem.c
+++ b/arch/nds32/mm/highmem.c
@@ -21,7 +21,7 @@ void *kmap_atomic_high(struct page *page)
idx = type + KM_TYPE_NR * smp_processor_id();
vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
- pte = (page_to_pfn(page) << PAGE_SHIFT) | (PAGE_KERNEL);
+ pte = (page_to_pfn(page) << PAGE_SHIFT) | (kmap_prot);
ptep = pte_offset_kernel(pmd_off_k(vaddr), vaddr);
set_pte(ptep, pte);
diff --git a/arch/xtensa/mm/highmem.c b/arch/xtensa/mm/highmem.c
index f57a7770eb08..8c58c4c37033 100644
--- a/arch/xtensa/mm/highmem.c
+++ b/arch/xtensa/mm/highmem.c
@@ -48,7 +48,7 @@ void *kmap_atomic_high(struct page *page)
#ifdef CONFIG_DEBUG_HIGHMEM
BUG_ON(!pte_none(*(kmap_pte + idx)));
#endif
- set_pte(kmap_pte + idx, mk_pte(page, PAGE_KERNEL_EXEC));
+ set_pte(kmap_pte + idx, mk_pte(page, kmap_prot));
return (void *)vaddr;
}
--
2.25.1
^ permalink raw reply related
* [PATCH V1 06/10] arch/kunmap_atomic: Consolidate duplicate code
From: ira.weiny @ 2020-04-30 20:38 UTC (permalink / raw)
To: linux-kernel, Andrew Morton, Christian Koenig, Huang Rui
Cc: Peter Zijlstra, Dave Hansen, dri-devel, James E.J. Bottomley,
Max Filippov, Paul Mackerras, H. Peter Anvin, sparclinux,
Ira Weiny, Thomas Gleixner, Helge Deller, x86, linux-csky,
Ingo Molnar, linux-snps-arc, linux-xtensa, Borislav Petkov,
Andy Lutomirski, Dan Williams, linux-arm-kernel, Chris Zankel,
Thomas Bogendoerfer, linux-parisc, linux-mips, linuxppc-dev,
David S. Miller
In-Reply-To: <20200430203845.582900-1-ira.weiny@intel.com>
From: Ira Weiny <ira.weiny@intel.com>
Every single architecture (including !CONFIG_HIGHMEM) calls...
pagefault_enable();
preempt_enable();
... before returning from __kunmap_atomic(). Lift this code into the
kunmap_atomic() macro.
While we are at it rename __kunmap_atomic() to kunmap_atomic_high() to
be consistent.
Signed-off-by: Ira Weiny <ira.weiny@intel.com>
---
Changes from V0:
rename __kunmap_atomic() to kunmap_atomic_high()
Fix mips issue
---
arch/arc/include/asm/highmem.h | 2 +-
arch/arc/mm/highmem.c | 7 ++-----
arch/arm/include/asm/highmem.h | 2 +-
arch/arm/mm/highmem.c | 6 ++----
arch/csky/include/asm/highmem.h | 2 +-
arch/csky/mm/highmem.c | 9 +++------
arch/microblaze/include/asm/highmem.h | 2 +-
arch/microblaze/mm/highmem.c | 6 ++----
arch/mips/include/asm/highmem.h | 2 +-
arch/mips/mm/cache.c | 4 ++--
arch/mips/mm/highmem.c | 6 ++----
arch/nds32/include/asm/highmem.h | 2 +-
arch/nds32/mm/highmem.c | 6 ++----
arch/parisc/include/asm/cacheflush.h | 4 +---
arch/powerpc/include/asm/highmem.h | 2 +-
arch/powerpc/mm/highmem.c | 6 ++----
arch/sparc/include/asm/highmem.h | 2 +-
arch/sparc/mm/highmem.c | 6 ++----
arch/x86/include/asm/highmem.h | 2 +-
arch/x86/mm/highmem_32.c | 7 ++-----
arch/xtensa/include/asm/highmem.h | 2 +-
arch/xtensa/mm/highmem.c | 7 ++-----
include/linux/highmem.h | 10 ++++++----
23 files changed, 40 insertions(+), 64 deletions(-)
diff --git a/arch/arc/include/asm/highmem.h b/arch/arc/include/asm/highmem.h
index 75bd0fa77fe2..e16531495620 100644
--- a/arch/arc/include/asm/highmem.h
+++ b/arch/arc/include/asm/highmem.h
@@ -31,7 +31,7 @@
#include <asm/cacheflush.h>
extern void *kmap_atomic_high(struct page *page);
-extern void __kunmap_atomic(void *kvaddr);
+extern void kunmap_atomic_high(void *kvaddr);
extern void kmap_init(void);
diff --git a/arch/arc/mm/highmem.c b/arch/arc/mm/highmem.c
index 0964b011c29f..5d3eab4ac0b0 100644
--- a/arch/arc/mm/highmem.c
+++ b/arch/arc/mm/highmem.c
@@ -65,7 +65,7 @@ void *kmap_atomic_high(struct page *page)
}
EXPORT_SYMBOL(kmap_atomic_high);
-void __kunmap_atomic(void *kv)
+void kunmap_atomic_high(void *kv)
{
unsigned long kvaddr = (unsigned long)kv;
@@ -87,11 +87,8 @@ void __kunmap_atomic(void *kv)
kmap_atomic_idx_pop();
}
-
- pagefault_enable();
- preempt_enable();
}
-EXPORT_SYMBOL(__kunmap_atomic);
+EXPORT_SYMBOL(kunmap_atomic_high);
static noinline pte_t * __init alloc_kmap_pgtable(unsigned long kvaddr)
{
diff --git a/arch/arm/include/asm/highmem.h b/arch/arm/include/asm/highmem.h
index 4edb6db3a5c8..a9d5e9bce1cc 100644
--- a/arch/arm/include/asm/highmem.h
+++ b/arch/arm/include/asm/highmem.h
@@ -61,7 +61,7 @@ static inline void *kmap_high_get(struct page *page)
*/
#ifdef CONFIG_HIGHMEM
extern void *kmap_atomic_high(struct page *page);
-extern void __kunmap_atomic(void *kvaddr);
+extern void kunmap_atomic_high(void *kvaddr);
extern void *kmap_atomic_pfn(unsigned long pfn);
#endif
diff --git a/arch/arm/mm/highmem.c b/arch/arm/mm/highmem.c
index 075fdc235091..ac8394655a6e 100644
--- a/arch/arm/mm/highmem.c
+++ b/arch/arm/mm/highmem.c
@@ -73,7 +73,7 @@ void *kmap_atomic_high(struct page *page)
}
EXPORT_SYMBOL(kmap_atomic_high);
-void __kunmap_atomic(void *kvaddr)
+void kunmap_atomic_high(void *kvaddr)
{
unsigned long vaddr = (unsigned long) kvaddr & PAGE_MASK;
int idx, type;
@@ -95,10 +95,8 @@ void __kunmap_atomic(void *kvaddr)
/* this address was obtained through kmap_high_get() */
kunmap_high(pte_page(pkmap_page_table[PKMAP_NR(vaddr)]));
}
- pagefault_enable();
- preempt_enable();
}
-EXPORT_SYMBOL(__kunmap_atomic);
+EXPORT_SYMBOL(kunmap_atomic_high);
void *kmap_atomic_pfn(unsigned long pfn)
{
diff --git a/arch/csky/include/asm/highmem.h b/arch/csky/include/asm/highmem.h
index 6807df1232f3..5bbbe59e60a9 100644
--- a/arch/csky/include/asm/highmem.h
+++ b/arch/csky/include/asm/highmem.h
@@ -33,7 +33,7 @@ extern pte_t *pkmap_page_table;
#define ARCH_HAS_KMAP_FLUSH_TLB
extern void kmap_flush_tlb(unsigned long addr);
extern void *kmap_atomic_high(struct page *page);
-extern void __kunmap_atomic(void *kvaddr);
+extern void kunmap_atomic_high(void *kvaddr);
extern void *kmap_atomic_pfn(unsigned long pfn);
extern struct page *kmap_atomic_to_page(void *ptr);
diff --git a/arch/csky/mm/highmem.c b/arch/csky/mm/highmem.c
index 63d74b47eee6..0aafbbbe651c 100644
--- a/arch/csky/mm/highmem.c
+++ b/arch/csky/mm/highmem.c
@@ -39,13 +39,13 @@ void *kmap_atomic_high(struct page *page)
}
EXPORT_SYMBOL(kmap_atomic_high);
-void __kunmap_atomic(void *kvaddr)
+void kunmap_atomic_high(void *kvaddr)
{
unsigned long vaddr = (unsigned long) kvaddr & PAGE_MASK;
int idx;
if (vaddr < FIXADDR_START)
- goto out;
+ return;
#ifdef CONFIG_DEBUG_HIGHMEM
idx = KM_TYPE_NR*smp_processor_id() + kmap_atomic_idx();
@@ -58,11 +58,8 @@ void __kunmap_atomic(void *kvaddr)
(void) idx; /* to kill a warning */
#endif
kmap_atomic_idx_pop();
-out:
- pagefault_enable();
- preempt_enable();
}
-EXPORT_SYMBOL(__kunmap_atomic);
+EXPORT_SYMBOL(kunmap_atomic_high);
/*
* This is the same as kmap_atomic() but can map memory that doesn't
diff --git a/arch/microblaze/include/asm/highmem.h b/arch/microblaze/include/asm/highmem.h
index fe4ad8bac9ae..5fc56b0107be 100644
--- a/arch/microblaze/include/asm/highmem.h
+++ b/arch/microblaze/include/asm/highmem.h
@@ -52,7 +52,7 @@ extern pte_t *pkmap_page_table;
#define PKMAP_ADDR(nr) (PKMAP_BASE + ((nr) << PAGE_SHIFT))
extern void *kmap_atomic_prot(struct page *page, pgprot_t prot);
-extern void __kunmap_atomic(void *kvaddr);
+extern void kunmap_atomic_high(void *kvaddr);
static inline void *kmap_atomic_high(struct page *page)
{
diff --git a/arch/microblaze/mm/highmem.c b/arch/microblaze/mm/highmem.c
index a14f356b055b..1026aeffe11a 100644
--- a/arch/microblaze/mm/highmem.c
+++ b/arch/microblaze/mm/highmem.c
@@ -51,7 +51,7 @@ void *kmap_atomic_prot(struct page *page, pgprot_t prot)
}
EXPORT_SYMBOL(kmap_atomic_prot);
-void __kunmap_atomic(void *kvaddr)
+void kunmap_atomic_high(void *kvaddr)
{
unsigned long vaddr = (unsigned long) kvaddr & PAGE_MASK;
int type;
@@ -77,7 +77,5 @@ void __kunmap_atomic(void *kvaddr)
local_flush_tlb_page(NULL, vaddr);
kmap_atomic_idx_pop();
- pagefault_enable();
- preempt_enable();
}
-EXPORT_SYMBOL(__kunmap_atomic);
+EXPORT_SYMBOL(kunmap_atomic_high);
diff --git a/arch/mips/include/asm/highmem.h b/arch/mips/include/asm/highmem.h
index a515bcf15d4b..d9f774bd4938 100644
--- a/arch/mips/include/asm/highmem.h
+++ b/arch/mips/include/asm/highmem.h
@@ -49,7 +49,7 @@ extern pte_t *pkmap_page_table;
#define ARCH_HAS_KMAP_FLUSH_TLB
extern void kmap_flush_tlb(unsigned long addr);
extern void *kmap_atomic_high(struct page *page);
-extern void __kunmap_atomic(void *kvaddr);
+extern void kunmap_atomic_high(void *kvaddr);
extern void *kmap_atomic_pfn(unsigned long pfn);
#define flush_cache_kmaps() BUG_ON(cpu_has_dc_aliases)
diff --git a/arch/mips/mm/cache.c b/arch/mips/mm/cache.c
index f015bb51fab0..1873c2a01fdb 100644
--- a/arch/mips/mm/cache.c
+++ b/arch/mips/mm/cache.c
@@ -104,7 +104,7 @@ void __flush_dcache_page(struct page *page)
flush_data_cache_page(addr);
if (PageHighMem(page))
- __kunmap_atomic((void *)addr);
+ kunmap_atomic((void *)addr);
}
EXPORT_SYMBOL(__flush_dcache_page);
@@ -147,7 +147,7 @@ void __update_cache(unsigned long address, pte_t pte)
flush_data_cache_page(addr);
if (PageHighMem(page))
- __kunmap_atomic((void *)addr);
+ kunmap_atomic((void *)addr);
ClearPageDcacheDirty(page);
}
diff --git a/arch/mips/mm/highmem.c b/arch/mips/mm/highmem.c
index 2bda56372995..155fbb107b35 100644
--- a/arch/mips/mm/highmem.c
+++ b/arch/mips/mm/highmem.c
@@ -36,7 +36,7 @@ void *kmap_atomic_high(struct page *page)
}
EXPORT_SYMBOL(kmap_atomic_high);
-void __kunmap_atomic(void *kvaddr)
+void kunmap_atomic_high(void *kvaddr)
{
unsigned long vaddr = (unsigned long) kvaddr & PAGE_MASK;
int type __maybe_unused;
@@ -63,10 +63,8 @@ void __kunmap_atomic(void *kvaddr)
}
#endif
kmap_atomic_idx_pop();
- pagefault_enable();
- preempt_enable();
}
-EXPORT_SYMBOL(__kunmap_atomic);
+EXPORT_SYMBOL(kunmap_atomic_high);
/*
* This is the same as kmap_atomic() but can map memory that doesn't
diff --git a/arch/nds32/include/asm/highmem.h b/arch/nds32/include/asm/highmem.h
index 28f5e7072c70..97648b678108 100644
--- a/arch/nds32/include/asm/highmem.h
+++ b/arch/nds32/include/asm/highmem.h
@@ -52,7 +52,7 @@ extern void kmap_init(void);
*/
#ifdef CONFIG_HIGHMEM
extern void *kmap_atomic_high(struct page *page);
-extern void __kunmap_atomic(void *kvaddr);
+extern void kunmap_atomic_high(void *kvaddr);
extern void *kmap_atomic_pfn(unsigned long pfn);
extern struct page *kmap_atomic_to_page(void *ptr);
#endif
diff --git a/arch/nds32/mm/highmem.c b/arch/nds32/mm/highmem.c
index f5f3a21460c4..f6e6915c0d31 100644
--- a/arch/nds32/mm/highmem.c
+++ b/arch/nds32/mm/highmem.c
@@ -34,7 +34,7 @@ void *kmap_atomic_high(struct page *page)
EXPORT_SYMBOL(kmap_atomic_high);
-void __kunmap_atomic(void *kvaddr)
+void kunmap_atomic_high(void *kvaddr)
{
if (kvaddr >= (void *)FIXADDR_START) {
unsigned long vaddr = (unsigned long)kvaddr;
@@ -45,8 +45,6 @@ void __kunmap_atomic(void *kvaddr)
ptep = pte_offset_kernel(pmd_off_k(vaddr), vaddr);
set_pte(ptep, 0);
}
- pagefault_enable();
- preempt_enable();
}
-EXPORT_SYMBOL(__kunmap_atomic);
+EXPORT_SYMBOL(kunmap_atomic_high);
diff --git a/arch/parisc/include/asm/cacheflush.h b/arch/parisc/include/asm/cacheflush.h
index 0c83644bfa5c..119c9a7681bc 100644
--- a/arch/parisc/include/asm/cacheflush.h
+++ b/arch/parisc/include/asm/cacheflush.h
@@ -122,11 +122,9 @@ static inline void *kmap_atomic(struct page *page)
return page_address(page);
}
-static inline void __kunmap_atomic(void *addr)
+static inline void kunmap_atomic_high(void *addr)
{
flush_kernel_dcache_page_addr(addr);
- pagefault_enable();
- preempt_enable();
}
#define kmap_atomic_prot(page, prot) kmap_atomic(page)
diff --git a/arch/powerpc/include/asm/highmem.h b/arch/powerpc/include/asm/highmem.h
index ac0efc2cf08a..1845fbd7ce61 100644
--- a/arch/powerpc/include/asm/highmem.h
+++ b/arch/powerpc/include/asm/highmem.h
@@ -60,7 +60,7 @@ extern pte_t *pkmap_page_table;
#define PKMAP_ADDR(nr) (PKMAP_BASE + ((nr) << PAGE_SHIFT))
extern void *kmap_atomic_prot(struct page *page, pgprot_t prot);
-extern void __kunmap_atomic(void *kvaddr);
+extern void kunmap_atomic_high(void *kvaddr);
static inline void *kmap_atomic_high(struct page *page)
{
diff --git a/arch/powerpc/mm/highmem.c b/arch/powerpc/mm/highmem.c
index f9558ef4b8fa..162958321e28 100644
--- a/arch/powerpc/mm/highmem.c
+++ b/arch/powerpc/mm/highmem.c
@@ -40,7 +40,7 @@ void *kmap_atomic_prot(struct page *page, pgprot_t prot)
}
EXPORT_SYMBOL(kmap_atomic_prot);
-void __kunmap_atomic(void *kvaddr)
+void kunmap_atomic_high(void *kvaddr)
{
unsigned long vaddr = (unsigned long) kvaddr & PAGE_MASK;
@@ -66,7 +66,5 @@ void __kunmap_atomic(void *kvaddr)
}
kmap_atomic_idx_pop();
- pagefault_enable();
- preempt_enable();
}
-EXPORT_SYMBOL(__kunmap_atomic);
+EXPORT_SYMBOL(kunmap_atomic_high);
diff --git a/arch/sparc/include/asm/highmem.h b/arch/sparc/include/asm/highmem.h
index c96a0603d821..94dd6e4c5fa4 100644
--- a/arch/sparc/include/asm/highmem.h
+++ b/arch/sparc/include/asm/highmem.h
@@ -51,7 +51,7 @@ void kmap_init(void) __init;
#define PKMAP_END (PKMAP_ADDR(LAST_PKMAP))
void *kmap_atomic_high(struct page *page);
-void __kunmap_atomic(void *kvaddr);
+void kunmap_atomic_high(void *kvaddr);
#define flush_cache_kmaps() flush_cache_all()
diff --git a/arch/sparc/mm/highmem.c b/arch/sparc/mm/highmem.c
index b53070ab6a31..469786bc430f 100644
--- a/arch/sparc/mm/highmem.c
+++ b/arch/sparc/mm/highmem.c
@@ -84,7 +84,7 @@ void *kmap_atomic_high(struct page *page)
}
EXPORT_SYMBOL(kmap_atomic_high);
-void __kunmap_atomic(void *kvaddr)
+void kunmap_atomic_high(void *kvaddr)
{
unsigned long vaddr = (unsigned long) kvaddr & PAGE_MASK;
int type;
@@ -126,7 +126,5 @@ void __kunmap_atomic(void *kvaddr)
#endif
kmap_atomic_idx_pop();
- pagefault_enable();
- preempt_enable();
}
-EXPORT_SYMBOL(__kunmap_atomic);
+EXPORT_SYMBOL(kunmap_atomic_high);
diff --git a/arch/x86/include/asm/highmem.h b/arch/x86/include/asm/highmem.h
index 72e154e17416..ff87aba96eee 100644
--- a/arch/x86/include/asm/highmem.h
+++ b/arch/x86/include/asm/highmem.h
@@ -63,7 +63,7 @@ void *kmap_atomic_high(struct page *page)
{
return kmap_atomic_prot(page, kmap_prot);
}
-void __kunmap_atomic(void *kvaddr);
+void kunmap_atomic_high(void *kvaddr);
void *kmap_atomic_pfn(unsigned long pfn);
void *kmap_atomic_prot_pfn(unsigned long pfn, pgprot_t prot);
diff --git a/arch/x86/mm/highmem_32.c b/arch/x86/mm/highmem_32.c
index 937d2cc40389..af07a6842743 100644
--- a/arch/x86/mm/highmem_32.c
+++ b/arch/x86/mm/highmem_32.c
@@ -30,7 +30,7 @@ void *kmap_atomic_pfn(unsigned long pfn)
}
EXPORT_SYMBOL_GPL(kmap_atomic_pfn);
-void __kunmap_atomic(void *kvaddr)
+void kunmap_atomic_high(void *kvaddr)
{
unsigned long vaddr = (unsigned long) kvaddr & PAGE_MASK;
@@ -60,11 +60,8 @@ void __kunmap_atomic(void *kvaddr)
BUG_ON(vaddr >= (unsigned long)high_memory);
}
#endif
-
- pagefault_enable();
- preempt_enable();
}
-EXPORT_SYMBOL(__kunmap_atomic);
+EXPORT_SYMBOL(kunmap_atomic_high);
void __init set_highmem_pages_init(void)
{
diff --git a/arch/xtensa/include/asm/highmem.h b/arch/xtensa/include/asm/highmem.h
index d4fb9f78ba32..a60a02dc68f6 100644
--- a/arch/xtensa/include/asm/highmem.h
+++ b/arch/xtensa/include/asm/highmem.h
@@ -69,7 +69,7 @@ static inline void flush_cache_kmaps(void)
}
void *kmap_atomic_high(struct page *page);
-void __kunmap_atomic(void *kvaddr);
+void kunmap_atomic_high(void *kvaddr);
void kmap_init(void);
diff --git a/arch/xtensa/mm/highmem.c b/arch/xtensa/mm/highmem.c
index 217f2ebaa298..f57a7770eb08 100644
--- a/arch/xtensa/mm/highmem.c
+++ b/arch/xtensa/mm/highmem.c
@@ -54,7 +54,7 @@ void *kmap_atomic_high(struct page *page)
}
EXPORT_SYMBOL(kmap_atomic_high);
-void __kunmap_atomic(void *kvaddr)
+void kunmap_atomic_high(void *kvaddr)
{
if (kvaddr >= (void *)FIXADDR_START &&
kvaddr < (void *)FIXADDR_TOP) {
@@ -73,11 +73,8 @@ void __kunmap_atomic(void *kvaddr)
kmap_atomic_idx_pop();
}
-
- pagefault_enable();
- preempt_enable();
}
-EXPORT_SYMBOL(__kunmap_atomic);
+EXPORT_SYMBOL(kunmap_atomic_high);
void __init kmap_init(void)
{
diff --git a/include/linux/highmem.h b/include/linux/highmem.h
index e0106b4f7dbb..601df07607a4 100644
--- a/include/linux/highmem.h
+++ b/include/linux/highmem.h
@@ -145,10 +145,10 @@ static inline void *kmap_atomic(struct page *page)
}
#define kmap_atomic_prot(page, prot) kmap_atomic(page)
-static inline void __kunmap_atomic(void *addr)
+static inline void kunmap_atomic_high(void *addr)
{
- pagefault_enable();
- preempt_enable();
+ /* Nothing to do in the CONFIG_HIGHMEM=n case as kunmap_atomic()
+ * handles re-enabling faults + preemption */
}
#define kmap_atomic_pfn(pfn) kmap_atomic(pfn_to_page(pfn))
@@ -198,7 +198,9 @@ static inline void kmap_atomic_idx_pop(void)
#define kunmap_atomic(addr) \
do { \
BUILD_BUG_ON(__same_type((addr), struct page *)); \
- __kunmap_atomic(addr); \
+ kunmap_atomic_high(addr); \
+ pagefault_enable(); \
+ preempt_enable(); \
} while (0)
--
2.25.1
^ permalink raw reply related
* Re: Don't initialise ports with no PHY
From: Darren Stevens @ 2020-04-30 20:45 UTC (permalink / raw)
To: Andrew Lunn; +Cc: oss, netdev, madalin.bacur, linuxppc-dev, chzigotzky
In-Reply-To: <20200425001021.GB1095011@lunn.ch>
Hello Andrew
On 25/04/2020, Andrew Lunn wrote:
> On Fri, Apr 24, 2020 at 11:29:38PM +0100, Darren Stevens wrote:
>> Since cbb961ca271e ("Use random MAC address when none is given")
>> Varisys Cyrus P5020 boards have been listing 5 ethernet ports instead of
>> the 2 the board has.This is because we were preventing the adding of the
>> unused ports by not suppling them a MAC address, which this patch now
>> supplies.
>>
>> Prevent them from appearing in the net devices list by checking for a
>> 'status="disabled"' entry during probe and skipping the port if we find
>> it.
>
> Hi Darren
>
> I'm surprised the core is probing a device which has status disabled.
> Are you sure this is the correct explanation?
You are correct, the core is detecting status="disabled". My mistake, the Hardware vendor's supplied dts always relied on only supplying active ports with an IP address, and it didn't occur to me to test that they could be disabled.
Sorry.
Regards
Darren
^ permalink raw reply
* Re: Don't initialise ports with no PHY
From: Darren Stevens @ 2020-04-30 21:32 UTC (permalink / raw)
To: Christian Zigotzky
Cc: Andrew Lunn, linuxppc-dev, mad skateman, R.T.Dickinson, netdev
In-Reply-To: <68BEA58A-DF1D-44B8-91DD-B90BAAB738BE@xenosoft.de>
[-- Attachment #1: Type: text/plain, Size: 178 bytes --]
AmigaOS...........: http://yam.ch/
Unix/MacOS/Windows: http://www.mozilla.com/thunderbird/
General information about MIME can be found at:
http://en.wikipedia.org/wiki/MIME
[-- Attachment #2: Type: text/plain, Size: 1153 bytes --]
Hello Christian
On 29/04/2020, Christian Zigotzky wrote:
>
>
>> On 29. Apr 2020, at 17:22, Andrew Lunn <andrew@lunn.ch> wrote:
>>
>> ?On Wed, Apr 29, 2020 at 03:55:28PM +0200, Christian Zigotzky wrote:
>>> Hi Andrew,
>>>
>>> You can find some dtb and source files in our kernel package.
>>>
>>> Download: http://www.xenosoft.de/linux-image-5.7-rc3-X1000_X5000.tar.gz
>>
>> I have the tarball. Are we talking about
>>
>>
linux-image-5.7-rc3-X1000_X5000/X5000_and_QEMU_e5500/dtbs/X5000_20/cyrus.eth.dtb
>
>> I don't see any status = "disabled"; in the blob. So i would expect
>> the driver to probe.
No, the vendor never added that to them.
> Yes, that's correct but maybe Darren uses another dtb file.
>
> @Darren
> Which dtb file do you use?
My current one attached, including updated cyrus_p5020.dts & p5020si-pre.dtsi
which I'm preparing patches for.
Christian, build an unmodified kernel, select board level reset or power off,
then both the GPIO drivers.
Then under LED Support: GPIO connected LED's and triggers -> disk activity
I think you still have a 5020 don't you? I'll look at 5040 later (I'll need
someone to test)
Regards
Darren
[-- Attachment #3: cyrus-dts.zip --]
[-- Type: application/zip, Size: 8289 bytes --]
^ permalink raw reply
* Re: [PATCH v2 2/3] mm/memory_hotplug: Introduce MHP_NO_FIRMWARE_MEMMAP
From: Andrew Morton @ 2020-04-30 22:24 UTC (permalink / raw)
To: David Hildenbrand
Cc: virtio-dev, linux-hyperv, Michal Hocko, Baoquan He, linux-acpi,
Wei Yang, linux-s390, linux-nvdimm, linux-kernel, virtualization,
linux-mm, Michael S . Tsirkin, Eric W. Biederman, Pankaj Gupta,
xen-devel, Michal Hocko, linuxppc-dev
In-Reply-To: <b28c9e02-8cf2-33ae-646b-fe50a185738e@redhat.com>
On Thu, 30 Apr 2020 20:43:39 +0200 David Hildenbrand <david@redhat.com> wrote:
> >
> > Why does the firmware map support hotplug entries?
>
> I assume:
>
> The firmware memmap was added primarily for x86-64 kexec (and still, is
> mostly used on x86-64 only IIRC). There, we had ACPI hotplug. When DIMMs
> get hotplugged on real HW, they get added to e820. Same applies to
> memory added via HyperV balloon (unless memory is unplugged via
> ballooning and you reboot ... the the e820 is changed as well). I assume
> we wanted to be able to reflect that, to make kexec look like a real reboot.
>
> This worked for a while. Then came dax/kmem. Now comes virtio-mem.
>
>
> But I assume only Andrew can enlighten us.
>
> @Andrew, any guidance here? Should we really add all memory to the
> firmware memmap, even if this contradicts with the existing
> documentation? (especially, if the actual firmware memmap will *not*
> contain that memory after a reboot)
For some reason that patch is misattributed - it was authored by
Shaohui Zheng <shaohui.zheng@intel.com>, who hasn't been heard from in
a decade. I looked through the email discussion from that time and I'm
not seeing anything useful. But I wasn't able to locate Dave Hansen's
review comments.
^ permalink raw reply
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