Daniel Axtens writes:
> Currently bitops-instrumented.h assumes that the architecture provides
> atomic, non-atomic and locking bitops (e.g. both set_bit and __set_bit).
> This is true on x86 and s390, but is not always true: there is a
> generic bitops/non-atomic.h header that provides generic
On Thu, Aug 29, 2019 at 10:38:10AM +0200, Christoph Hellwig wrote:
> On Thu, Aug 22, 2019 at 03:56:14PM +0530, Bharata B Rao wrote:
> > +/*
> > + * Bits 60:56 in the rmap entry will be used to identify the
> > + * different uses/functions of rmap.
> > + */
> > +#define KVMPPC_RMAP_DEVM_PFN
Xmon should be either fully or partially disabled depending on the
kernel lockdown state.
Put xmon into read-only mode for lockdown=integrity and prevent user
entry into xmon when lockdown=confidentiality. Xmon checks the lockdown
state on every attempted entry:
(1) during early xmon'ing
(2)
Xmon should be either fully or partially disabled depending on the
kernel lockdown state.
Put xmon into read-only mode for lockdown=integrity and completely
disable xmon when lockdown=confidentiality. Since this can occur
dynamically, there may be pre-existing, active breakpoints in xmon when
Read-only mode should not prevent listing and clearing any active
breakpoints.
Tested-by: Daniel Axtens
Reviewed-by: Daniel Axtens
Signed-off-by: Christopher M. Riedl
---
arch/powerpc/xmon/xmon.c | 16 +++-
1 file changed, 11 insertions(+), 5 deletions(-)
diff --git
On Wed, 2019-08-28 at 08:07 +0200, Thomas Huth wrote:
> The FSF does not reside in "675 Mass Ave, Cambridge" anymore...
> let's simply use proper SPDX identifiers instead.
>
> Signed-off-by: Thomas Huth
Acked-by: Russell Currey
On Wed, 2019-08-28 at 15:54 +0200, Christophe Leroy wrote:
> Any plan to getting this applied soon ?
Hey Christophe,
I'm still working on it. Had to rework it for a few reasons, and it
exposed a bug somewhere else. Hope to have another version out soon.
- Russell
>
> Christophe
Subject: Singaporean Mr. Teo En Ming's Refugee Seeking Attempts
In reverse chronological order:
[1] Petition to the Government of Taiwan for Refugee Status, 5th August 2019
Monday
Photo #1: At the building of the National Immigration Agency, Ministry of the
Interior, Taipei, Taiwan, 5th
The argument lists for the INT_HANDLER macro are getting a bit
unwieldy. Use keyword parameters with default values to shorten them.
Signed-off-by: Michael Ellerman
---
arch/powerpc/kernel/exceptions-64s.S | 120 +--
1 file changed, 60 insertions(+), 60 deletions(-)
Provide the current number of vmalloc shadow pages in
/sys/kernel/debug/kasan_vmalloc/shadow_pages.
Signed-off-by: Daniel Axtens
---
Merging this is probably overkill, but I leave it to the discretion
of the broader community.
On v4 (no dynamic freeing), I saw the following approximate
In the case where KASAN directly allocates memory to back vmalloc
space, don't map the early shadow page over it.
We prepopulate pgds/p4ds for the range that would otherwise be empty.
This is required to get it synced to hardware on boot, allowing the
lower levels of the page tables to be filled
Supporting VMAP_STACK with KASAN_VMALLOC is straightforward:
- clear the shadow region of vmapped stacks when swapping them in
- tweak Kconfig to allow VMAP_STACK to be turned on with KASAN
Reviewed-by: Dmitry Vyukov
Signed-off-by: Daniel Axtens
---
arch/Kconfig | 9 +
Test kasan vmalloc support by adding a new test to the module.
Signed-off-by: Daniel Axtens
--
v5: split out per Christophe Leroy
---
lib/test_kasan.c | 26 ++
1 file changed, 26 insertions(+)
diff --git a/lib/test_kasan.c b/lib/test_kasan.c
index
Hook into vmalloc and vmap, and dynamically allocate real shadow
memory to back the mappings.
Most mappings in vmalloc space are small, requiring less than a full
page of shadow space. Allocating a full shadow page per mapping would
therefore be wasteful. Furthermore, to ensure that different
Currently, vmalloc space is backed by the early shadow page. This
means that kasan is incompatible with VMAP_STACK.
This series provides a mechanism to back vmalloc space with real,
dynamically allocated memory. I have only wired up x86, because that's
the only currently supported arch I can work
Commit 36f1031c51a2 ("ibmvnic: Do not process reset during or after
device removal") made the change to exit reset if the driver has been
removed, but does not free reset work items of the adapter from queue.
Ensure all reset work items are freed when breaking out of the loop early.
Fixes:
Nicholas Piggin's on August 30, 2019 1:49 am:
>> This doesn't do what you think, or what you want, or what you think you
>> want ;-)
>
> After reading gcc docs from gcc 4 to 9, I think it does.
>
> We want this to apply to all functions in the compilaition unit. It's
> fine to use the regs
Building callchain.c with !COMPAT proved quite ugly with all the
defines. Splitting out the 32bit and 64bit parts looks better.
Also rewrite current_is_64bit as common function. No other code change
intended.
Signed-off-by: Michal Suchanek
---
arch/powerpc/perf/Makefile | 4 +
On bigendian ppc64 it is common to have 32bit legacy binaries but much
less so on littleendian.
Signed-off-by: Michal Suchanek
---
v3: make configurable
---
arch/powerpc/Kconfig | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig
There are numerous references to 32bit functions in generic and 64bit
code so ifdef them out.
Signed-off-by: Michal Suchanek
---
v2:
- fix 32bit ifdef condition in signal.c
- simplify the compat ifdef condition in vdso.c - 64bit is redundant
- simplify the compat ifdef condition in callchain.c -
These functions are required for 64bit as well.
Signed-off-by: Michal Suchanek
---
arch/powerpc/kernel/signal.c| 141
arch/powerpc/kernel/signal_32.c | 140 ---
2 files changed, 141 insertions(+), 140 deletions(-)
diff --git
The llseek syscall is not built in fs/read_write.c when !64bit && !COMPAT
With the syscall marked as common in syscall.tbl build fails in this
case.
The llseek inteface does not make sense on 64bit and it is explicitly
described as 32bit interface. Use on 64bit is not well-defined so just
drop it
Less code means less bugs so add a knob to skip the compat stuff.
This is tested on ppc64le top of
https://patchwork.ozlabs.org/cover/1153556/
Changes in v2: saner CONFIG_COMPAT ifdefs
Changes in v3:
- change llseek to 32bit instead of builing it unconditionally in fs
- clanup the makefile
On Thu, 29 Aug 2019 15:34:08 -0500
Uma Krishnan wrote:
> Below commit queued up for 5.4 includes these changes.
>
> commit 657bd277c162580674ddb86a90c4aeb62639bff5
> Author: Gustavo A. R. Silva
> Date: Sun Jul 28 19:21:19 2019 -0500
>
> Thanks,
> Uma Krishnan
Works for me as well.
Thanks
Below commit queued up for 5.4 includes these changes.
commit 657bd277c162580674ddb86a90c4aeb62639bff5
Author: Gustavo A. R. Silva
Date: Sun Jul 28 19:21:19 2019 -0500
Thanks,
Uma Krishnan
> On Aug 29, 2019, at 7:32 AM, Michal Suchanek wrote:
>
> Add fallthrough comments where missing.
>
Below commit queued up for 5.4 includes these changes.commit 657bd277c162580674ddb86a90c4aeb62639bff5Author: Gustavo A. R. Silva Date: Sun Jul 28 19:21:19 2019 -0500Thanks,Uma KrishnanOn Aug 29, 2019, at 7:32 AM, Michal Suchanek wrote:Add fallthrough comments where missing.Signed-off-by: Michal
The following commit has been merged into the perf/core branch of tip:
Commit-ID: 0a56e0603fa13af08816d673f6f71b68cda2fb2e
Gitweb:
https://git.kernel.org/tip/0a56e0603fa13af08816d673f6f71b68cda2fb2e
Author:Naveen N. Rao
AuthorDate:Tue, 27 Aug 2019 12:44:58 +05:30
On 8/27/19 8:18 AM, Andy Shevchenko wrote:
Kernel-doc validator complains:
aer.c:207: warning: Function parameter or member 'str' not described in
'pcie_ecrc_get_policy'
aer.c:1209: warning: Function parameter or member 'irq' not described in
'aer_isr'
aer.c:1209: warning: Function
On 8/27/19 8:18 AM, Andy Shevchenko wrote:
This simplifies and standardizes slot manipulation code
by using for_each_set_bit() library function.
Signed-off-by: Andy Shevchenko
Reviewed-by: Kuppuswamy Sathyanarayanan
---
drivers/pci/pcie/aer.c | 19 ---
1 file changed, 8
On Thu, 29 Aug 2019 16:32:50 +0200
Arnd Bergmann wrote:
> On Thu, Aug 29, 2019 at 4:19 PM Michal Suchánek wrote:
> > On Thu, 29 Aug 2019 14:57:39 +0200 Arnd Bergmann wrote:
> > > On Thu, Aug 29, 2019 at 2:37 PM Michal Suchánek
> > > wrote:
> > > > On Thu, 29 Aug 2019 14:19:46 +0200 Arnd
Bharata B Rao [bhar...@linux.ibm.com] wrote:
> On Wed, Aug 28, 2019 at 08:02:19PM -0700, Sukadev Bhattiprolu wrote:
> > Some minor comments/questions below. Overall, the patches look
> > fine to me.
> >
> > > +#include
> > > +#include
> > > +#include
> > > +#include
> > > +
> > > +static
On Thu, 29 Aug 2019 19:40:55 +0200
Christophe Leroy wrote:
> Le 29/08/2019 à 12:23, Michal Suchanek a écrit :
> > There are numerous references to 32bit functions in generic and 64bit
> > code so ifdef them out.
> >
> > Signed-off-by: Michal Suchanek
> > ---
> > v2:
> > - fix 32bit ifdef
From: Thomas Falcon
[ Upstream commit 66cf4710b23ab2adda11155684a2c8826f4fe732 ]
The ibm,mac-address-filters property defines the maximum number of
addresses the hypervisor's multicast filter list can support. It is
encoded as a big-endian integer in the OF device tree, but the virtual
ethernet
From: Thomas Falcon
[ Upstream commit 66cf4710b23ab2adda11155684a2c8826f4fe732 ]
The ibm,mac-address-filters property defines the maximum number of
addresses the hypervisor's multicast filter list can support. It is
encoded as a big-endian integer in the OF device tree, but the virtual
ethernet
From: Thomas Falcon
[ Upstream commit 66cf4710b23ab2adda11155684a2c8826f4fe732 ]
The ibm,mac-address-filters property defines the maximum number of
addresses the hypervisor's multicast filter list can support. It is
encoded as a big-endian integer in the OF device tree, but the virtual
ethernet
From: Thomas Falcon
[ Upstream commit 80f0fe0934cd3daa13a5e4d48a103f469115b160 ]
There's no need to wait until a completion is received to unmap
TX descriptor buffers that have been passed to the hypervisor.
Instead unmap it when the hypervisor call has completed. This patch
avoids the
From: Thomas Falcon
[ Upstream commit 66cf4710b23ab2adda11155684a2c8826f4fe732 ]
The ibm,mac-address-filters property defines the maximum number of
addresses the hypervisor's multicast filter list can support. It is
encoded as a big-endian integer in the OF device tree, but the virtual
ethernet
From: Thomas Falcon
[ Upstream commit 80f0fe0934cd3daa13a5e4d48a103f469115b160 ]
There's no need to wait until a completion is received to unmap
TX descriptor buffers that have been passed to the hypervisor.
Instead unmap it when the hypervisor call has completed. This patch
avoids the
From: Thomas Falcon
[ Upstream commit 66cf4710b23ab2adda11155684a2c8826f4fe732 ]
The ibm,mac-address-filters property defines the maximum number of
addresses the hypervisor's multicast filter list can support. It is
encoded as a big-endian integer in the OF device tree, but the virtual
ethernet
Le 29/08/2019 à 12:23, Michal Suchanek a écrit :
There are numerous references to 32bit functions in generic and 64bit
code so ifdef them out.
Signed-off-by: Michal Suchanek
---
v2:
- fix 32bit ifdef condition in signal.c
- simplify the compat ifdef condition in vdso.c - 64bit is redundant
On 8/28/19 5:39 AM, Mimi Zohar wrote:
> Detect and allow appended signatures.
Ran the tests successfully on x86_64 QEMU
Reviewed-by: Jordan Hand
Tested-by: Jordan Hand
From: Claudio Carvalho
Add the PowerPC name and the PPC_ELFNOTE_CAPABILITIES type in the
kernel binary ELF note. This type is a bitmap that can be used to
advertise kernel capabilities to userland.
This patch also defines PPCCAP_ULTRAVISOR_BIT as being the bit zero.
Suggested-by: Paul
The ELF note documentation describes the types and descriptors to be
used with the PowerPC namespace.
Signed-off-by: Maxiwell S. Garcia
Signed-off-by: Claudio Carvalho
---
Documentation/powerpc/elfnote.rst | 42 +++
1 file changed, 42 insertions(+)
create mode
The first patch adds the PowerPC name and the PPC_ELFNOTE_CAPABILITIES
type in the kernel binary ELF note. This type is a bitmap that can be
used to advertise kernel capabilities to userland.
The second patch adds a new documentation file to detail the
ELF Note PowerPC namespace and the
Segher Boessenkool's on August 29, 2019 9:51 pm:
> On Thu, Aug 29, 2019 at 07:38:01PM +1000, Nicholas Piggin wrote:
>> So... this actually seems to work. Haven't booted it, but the compiler
>> seems to do what we want.
>
> From the GCC manual:
>
>After defining a global register variable,
I've applied this to the dma-mapping for-next tree now.
If there are any issues with the parisc patch I'll happily take
incremental patches.
I've pulled this into the dma-mapping for-next tree now.
From: "Naveen N. Rao"
Copy over powerpc syscall.tbl to grab changes from the below commits:
commit cee3536d24a1 ("powerpc: Wire up clone3 syscall")
commit 1a271a68e030 ("arch: mark syscall number 435 reserved for clone3")
commit 7615d9e1780e ("arch: wire-up pidfd_open()")
commit
On Thu, Aug 29, 2019 at 4:19 PM Michal Suchánek wrote:
> On Thu, 29 Aug 2019 14:57:39 +0200 Arnd Bergmann wrote:
> > On Thu, Aug 29, 2019 at 2:37 PM Michal Suchánek wrote:
> > > On Thu, 29 Aug 2019 14:19:46 +0200 Arnd Bergmann wrote:
> > > > On Thu, Aug 29, 2019 at 12:23 PM Michal Suchanek
>
On Thu, 29 Aug 2019 14:57:39 +0200
Arnd Bergmann wrote:
> On Thu, Aug 29, 2019 at 2:37 PM Michal Suchánek wrote:
> > On Thu, 29 Aug 2019 14:19:46 +0200 Arnd Bergmann wrote:
> > > On Thu, Aug 29, 2019 at 12:23 PM Michal Suchanek
> > > wrote:
> > > In particular, I don't see why you single
From: Nathan Chancellor
> Sent: 28 August 2019 19:45
...
> However, I think that -fno-builtin-* would be appropriate here because
> we are providing our own setjmp implementation, meaning clang should not
> be trying to do anything with the builtin implementation like building a
> declaration for
On 2019-08-29, Rasmus Villemoes wrote:
> On 29/08/2019 14.15, Aleksa Sarai wrote:
> > On 2019-08-24, Daniel Colascione wrote:
>
> >> Why pad the structure when new functionality (perhaps accommodated via
> >> a larger structure) could be signaled by passing a new flag? Adding
> >> reserved
On 29/08/2019 14.15, Aleksa Sarai wrote:
> On 2019-08-24, Daniel Colascione wrote:
>> Why pad the structure when new functionality (perhaps accommodated via
>> a larger structure) could be signaled by passing a new flag? Adding
>> reserved fields to a structure with a size embedded in the ABI
Le 29/08/2019 à 14:14, Michael Ellerman a écrit :
Christophe Leroy writes:
DSISR has a bit to tell if the fault is due to a read or a write.
Except some CPUs don't have a DSISR?
Which is why we have page_fault_is_write() that's used in
__do_page_fault().
Or is that old cruft?
I see eg.
On Thu, Aug 29, 2019 at 2:37 PM Michal Suchánek wrote:
> On Thu, 29 Aug 2019 14:19:46 +0200 Arnd Bergmann wrote:
> > On Thu, Aug 29, 2019 at 12:23 PM Michal Suchanek wrote:
> > In particular, I don't see why you single out llseek here, but leave other
> > syscalls that are not needed on 64-bit
> On August 29, 2019 at 1:40 AM Daniel Axtens wrote:
>
>
> Hi Chris,
>
> > Read-only mode should not prevent listing and clearing any active
> > breakpoints.
>
> I tested this and it works for me:
>
> Tested-by: Daniel Axtens
>
> > + if (xmon_is_ro || !scanhex()) {
>
> It took
On Thu, 29 Aug 2019 14:19:46 +0200
Arnd Bergmann wrote:
> On Thu, Aug 29, 2019 at 12:23 PM Michal Suchanek wrote:
> >
> > Fixes: aff850393200 ("powerpc: add system call table generation support")
>
> This patch needs a proper explanation. The Fixes tag doesn't seem right
> here, since ppc64
> On August 29, 2019 at 2:43 AM Daniel Axtens wrote:
>
>
> Hi,
>
> > Xmon should be either fully or partially disabled depending on the
> > kernel lockdown state.
>
> I've been kicking the tyres of this, and it seems to work well:
>
> Tested-by: Daniel Axtens
>
Thank you for taking the
Add fallthrough comments where missing.
Signed-off-by: Michal Suchanek
---
drivers/scsi/cxlflash/main.c | 8
1 file changed, 8 insertions(+)
diff --git a/drivers/scsi/cxlflash/main.c b/drivers/scsi/cxlflash/main.c
index b1f4724efde2..f402fa9a7bec 100644
---
Add fallthrough comments where they are missing.
Signed-off-by: Michal Suchanek
---
drivers/scsi/ibmvscsi/ibmvfc.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/drivers/scsi/ibmvscsi/ibmvfc.c b/drivers/scsi/ibmvscsi/ibmvfc.c
index 8cdbac076a1b..2a06a5b4d3a5 100644
---
Add fallthrough comments where they are missing.
Signed-off-by: Michal Suchanek
---
drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c
b/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c
index
Build with recent compiler results in numerous fallthrough warnings.
The code looks correct to me so just adding the comments to quiet gcc.
Please review if the fallthrough was really intentional.
Thanks
Michal
Michal Suchanek (3):
scsi: cxlflash: Fix fallthrough warnings.
scsi: ibmvfc:
On Thu, Aug 29, 2019 at 12:23 PM Michal Suchanek wrote:
>
> Fixes: aff850393200 ("powerpc: add system call table generation support")
This patch needs a proper explanation. The Fixes tag doesn't seem right
here, since ppc64 has had llseek since the start in 2002 commit 3939e37587e7
("Add ppc64
On 2019-08-24, Daniel Colascione wrote:
> On Mon, Aug 19, 2019 at 8:37 PM Aleksa Sarai wrote:
> >
> > The most obvious syscall to add support for the new LOOKUP_* scoping
> > flags would be openat(2). However, there are a few reasons why this is
> > not the best course of action:
> >
> > * The
Christophe Leroy writes:
> DSISR has a bit to tell if the fault is due to a read or a write.
Except some CPUs don't have a DSISR?
Which is why we have page_fault_is_write() that's used in
__do_page_fault().
Or is that old cruft?
I see eg. in head_40x.S we pass r5=0 for error code, and we
On Thu, Aug 29, 2019 at 07:38:01PM +1000, Nicholas Piggin wrote:
> So... this actually seems to work. Haven't booted it, but the compiler
> seems to do what we want.
>From the GCC manual:
After defining a global register variable, for the current compilation
unit:
* If the register is
> -Original Message-
> From: Scott Wood
> Sent: Wednesday, August 28, 2019 7:19 AM
> To: Valentin Longchamp ; Madalin-cristian Bucur
>
> Cc: linuxppc-dev@lists.ozlabs.org; ga...@kernel.crashing.org;
> net...@vger.kernel.org
> Subject: Re: [PATCH] powerpc/kmcent2: update the ethernet
From: Christophe Leroy
[ Upstream commit 9c4e4c90ec24652921e31e9551fcaedc26eec86d ]
Otherwise, the following warning is encountered:
WARNING: vmlinux.o(.text+0x3dc6): Section mismatch in reference from the
variable start_here_multiplatform to the function .init.text:.early_setup()
The
From: Christophe Leroy
[ Upstream commit 9c4e4c90ec24652921e31e9551fcaedc26eec86d ]
Otherwise, the following warning is encountered:
WARNING: vmlinux.o(.text+0x3dc6): Section mismatch in reference from the
variable start_here_multiplatform to the function .init.text:.early_setup()
The
Nicholas Piggin's on August 29, 2019 7:38 pm:
> Christophe Leroy's on August 28, 2019 7:55 pm:
>>
>>
>> Le 28/08/2019 à 11:49, Nicholas Piggin a écrit :
>>> Christophe Leroy's on August 28, 2019 7:06 pm:
Le 27/08/2019 à 15:55, Nicholas Piggin a écrit :
> Accounted for some
On Thu, 29 Aug 2019 12:23:42 +0200
Michal Suchanek wrote:
> There are numerous references to 32bit functions in generic and 64bit
> code so ifdef them out.
>
> Signed-off-by: Michal Suchanek
> ---
> v2:
> - fix 32bit ifdef condition in signal.c
> - simplify the compat ifdef condition in vdso.c
On bigendian ppc64 it is common to have 32bit legacy binaries but much
less so on littleendian.
Signed-off-by: Michal Suchanek
---
v3: make configurable
---
arch/powerpc/Kconfig | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig
There are numerous references to 32bit functions in generic and 64bit
code so ifdef them out.
Signed-off-by: Michal Suchanek
---
v2:
- fix 32bit ifdef condition in signal.c
- simplify the compat ifdef condition in vdso.c - 64bit is redundant
- simplify the compat ifdef condition in callchain.c -
These functions are required for 64bit as well.
Signed-off-by: Michal Suchanek
---
arch/powerpc/kernel/signal.c| 141
arch/powerpc/kernel/signal_32.c | 140 ---
2 files changed, 141 insertions(+), 140 deletions(-)
diff --git
Fixes: aff850393200 ("powerpc: add system call table generation support")
Signed-off-by: Michal Suchanek
---
arch/powerpc/kernel/syscalls/syscall.tbl | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/arch/powerpc/kernel/syscalls/syscall.tbl
Less code means less bugs so add a knob to skip the compat stuff.
This is tested on ppc64le top of
https://patchwork.ozlabs.org/cover/1153556/
Changes in v2: saner CONFIG_COMPAT ifdefs
Changes in v3:
- change llseek to 32bit instead of builing it unconditionally in fs
- clanup the makefile
On Wed, 28 Aug 2019 23:22:00 -0700
Christoph Hellwig wrote:
> On Wed, Aug 28, 2019 at 06:15:40PM +0200, Michal Suchánek wrote:
> > On Wed, 28 Aug 2019 08:15:52 -0700
> > Christoph Hellwig wrote:
> >
> > > On Tue, Aug 27, 2019 at 10:21:06PM +0200, Michal Suchanek wrote:
> > > > 64bit
On Wed, 28 Aug 2019 23:46:24 -0700
Christoph Hellwig wrote:
> On Wed, Aug 28, 2019 at 06:43:50PM +0200, Michal Suchanek wrote:
> > +ifdef CONFIG_COMPAT
> > +obj-y += sys_ppc32.o ptrace32.o signal_32.o
> > +endif
>
> This should be:
>
> obj-$(CONFIG_COMPAT)
Christoph Hellwig writes:
> On Wed, Aug 28, 2019 at 11:34:09PM +1000, Michael Ellerman wrote:
>> Christoph Hellwig writes:
>> > Michael,
>> >
>> > do oyu plan to pick this up? Otherwise I'd love to pick it up through
>> > the dma-mapping tree as that would avoid one trivial conflict with
>> >
Christophe Leroy's on August 28, 2019 7:55 pm:
>
>
> Le 28/08/2019 à 11:49, Nicholas Piggin a écrit :
>> Christophe Leroy's on August 28, 2019 7:06 pm:
>>>
>>>
>>> Le 27/08/2019 à 15:55, Nicholas Piggin a écrit :
Accounted for some feedback.
Nicholas Piggin (4):
powerpc:
> -Original Message-
> From: Rob Herring
> Sent: 2019年8月28日 6:26
> To: Xiaowei Bao
> Cc: bhelg...@google.com; mark.rutl...@arm.com; shawn...@kernel.org;
> Leo Li ; kis...@ti.com; lorenzo.pieral...@arm.co;
> a...@arndb.de; gre...@linuxfoundation.org; M.h. Lian
> ; Mingkai Hu ; Roy Zang
On Thu, Aug 29, 2019 at 6:21 PM Aneesh Kumar K.V
wrote:
>
> On 8/29/19 1:29 PM, Oliver O'Halloran wrote:
> > On Thu, Aug 29, 2019 at 4:34 PM Aneesh Kumar K.V
> > wrote:
> >>
> >> Right now we force an unbind of SCM memory at drcindex on H_OVERLAP error.
> >> This really slows down operations
At the moment updates in a TCE table are made by iommu_table_ops::exchange
which update one TCE and invalidates an entry in the PHB/NPU TCE cache
via set of registers called "TCE Kill" (hence the naming).
Writing a TCE is a simple xchg() but invalidating the TCE cache is
a relatively expensive
As now we have xchg_no_kill/tce_kill, these are not used anymore so
remove them.
Signed-off-by: Alexey Kardashevskiy
---
arch/powerpc/include/asm/iommu.h | 10 -
arch/powerpc/kernel/iommu.c | 26 +---
arch/powerpc/platforms/powernv/pci-ioda.c | 50
This is the last implementation of iommu_table_ops::exchange() which
we are about to remove.
This implements xchg_no_kill() for pseries. Since it is paravirtual
platform, the hypervisor does TCE invalidations and we do not have
to deal with it here, hence no tce_kill() hook.
Signed-off-by:
Invalidating a TCE cache entry for each updated TCE is quite expensive.
This makes use of the new iommu_table_ops::xchg_no_kill()/tce_kill()
callbacks to bring down the time spent in mapping a huge guest DMA window.
Signed-off-by: Alexey Kardashevskiy
---
drivers/vfio/vfio_iommu_spapr_tce.c |
Invalidating a TCE cache entry for each updated TCE is quite expensive.
This makes use of the new iommu_table_ops::xchg_no_kill()/tce_kill()
callbacks to bring down the time spent in mapping a huge guest DMA window;
roughly 20s to 10s for each guest's 100GB of DMA space.
Signed-off-by: Alexey
So far TCE cache updates (IOMMU translation cache on POWER8/9
PHB/NPU units) were barely noticeable; however with 100+GB guests
we now see RCU stall warnings in guests because we spend too much
time in the host system firmware which does actual TCE cache
updates, hence this patchset.
This is a
On Thu, Aug 29, 2019 at 10:38 AM Christophe Leroy
wrote:
> Le 29/08/2019 à 10:01, Arnd Bergmann a écrit :
> > On Thu, Aug 29, 2019 at 8:46 AM Christoph Hellwig
> > wrote:
> >
> >>> @@ -277,7 +277,7 @@ static void do_signal(struct task_struct *tsk)
> >>>
> >>>rseq_signal_deliver(,
Most 8xx registers have specific names, so just include
reg_8xx.h all the time in reg.h in order to have them defined
even when CONFIG_PPC_8xx is not selected. This will avoid
the need for #ifdefs in C code.
Guard SPRN_ICTRL in an #ifdef CONFIG_PPC_8xx as this register
has same name but different
SET_MSR_EE() is just use in this file and doesn't provide
any added value compared to mtmsr(). Drop it.
Add a wrtee() inline function to use wrtee/wrteei insn.
Replace #ifdefs by IS_ENABLED()
Signed-off-by: Christophe Leroy
---
v2: Changed wrtee()/wrteei() to a single wrtee() inline which
The existing code uses bunch of hardcoded values from the PCI Bus Binding
to IEEE Std 1275 spec; and it does so in quite non-obvious way.
This defines fields from the cell#0 of the "reg" property of a PCI device
and uses them for parsing.
This should cause no behavioral change.
Signed-off-by:
On Thu, Aug 22, 2019 at 03:56:17PM +0530, Bharata B Rao wrote:
> + /*
> + * TODO: Handle KVM_MR_MOVE
> + */
> + if (change == KVM_MR_CREATE) {
> + uv_register_mem_slot(kvm->arch.lpid,
> + new->base_gfn << PAGE_SHIFT,
> +
Le 29/08/2019 à 10:01, Arnd Bergmann a écrit :
On Thu, Aug 29, 2019 at 8:46 AM Christoph Hellwig wrote:
@@ -277,7 +277,7 @@ static void do_signal(struct task_struct *tsk)
rseq_signal_deliver(, tsk->thread.regs);
- if (is32) {
+ if ((IS_ENABLED(CONFIG_PPC32) ||
On Thu, Aug 22, 2019 at 03:56:14PM +0530, Bharata B Rao wrote:
> +/*
> + * Bits 60:56 in the rmap entry will be used to identify the
> + * different uses/functions of rmap.
> + */
> +#define KVMPPC_RMAP_DEVM_PFN (0x2ULL << 56)
How did you come up with this specific value?
> +
> +static inline
On Wed, Aug 28, 2019 at 03:16:19PM -0700, Nick Desaulniers wrote:
> That's a good reason IMO. IIRC, the -fno-builtin-* flags don't warn
> if * is some unrecognized value, so -fno-builtin-setjmp may not
> actually do anything, and you may need to scan the source (of clang or
> llvm).
On 8/29/19 1:29 PM, Oliver O'Halloran wrote:
On Thu, Aug 29, 2019 at 4:34 PM Aneesh Kumar K.V
wrote:
Right now we force an unbind of SCM memory at drcindex on H_OVERLAP error.
This really slows down operations like kexec where we get the H_OVERLAP
error because we don't go through a full
On Thu, Aug 29, 2019 at 8:46 AM Christoph Hellwig wrote:
> > @@ -277,7 +277,7 @@ static void do_signal(struct task_struct *tsk)
> >
> > rseq_signal_deliver(, tsk->thread.regs);
> >
> > - if (is32) {
> > + if ((IS_ENABLED(CONFIG_PPC32) || IS_ENABLED(CONFIG_COMPAT)) && is32) {
>
> I
On Thu, Aug 29, 2019 at 4:34 PM Aneesh Kumar K.V
wrote:
>
> Right now we force an unbind of SCM memory at drcindex on H_OVERLAP error.
> This really slows down operations like kexec where we get the H_OVERLAP
> error because we don't go through a full hypervisor re init.
Maybe we should be
On Wed, Aug 28, 2019 at 08:05:52PM -0700, Sukadev Bhattiprolu wrote:
> > - After the guest becomes secure, when we handle a page fault of a page
> > belonging to SVM in HV, send that page to UV via UV_PAGE_IN.
> > - Whenever a page is unmapped on the HV side, inform UV via UV_PAGE_INVAL.
> > -
1 - 100 of 113 matches
Mail list logo
