to the current kernel interfaces such as
using const for struct device_type and changing the platform remove
callback signature.
Thank you,
Ira Weiny
---
The following changes since commit ed30a4a51bb196781c8058073ea720133a65596f:
Linux 6.9-rc5 (2024-04-21 12:35:54 -0700)
are available in the Git
Jonathan Cameron wrote:
> On Wed, 14 Feb 2024 17:33:18 -0500
> Steven Rostedt wrote:
>
> > On Wed, 14 Feb 2024 14:19:19 -0800
> > Ira Weiny wrote:
> >
> > > > > Jonathan Cameron wrote:
> > > > >
> > > >
Ira Weiny wrote:
> Jonathan Cameron wrote:
> > On Wed, 14 Feb 2024 10:23:10 -0500
> > Steven Rostedt wrote:
> >
> > > On Wed, 14 Feb 2024 12:11:53 +
> > > Jonathan Cameron wrote:
> > >
> > > > So I'm thinking this
Jonathan Cameron wrote:
> On Wed, 14 Feb 2024 10:23:10 -0500
> Steven Rostedt wrote:
>
> > On Wed, 14 Feb 2024 12:11:53 +
> > Jonathan Cameron wrote:
> >
> > > So I'm thinking this is a won't fix - wait for the printk rework to land
> > > and
> > > assume this will be resolved as well?
Jonathan Cameron wrote:
> On Wed, 14 Feb 2024 10:23:10 -0500
> Steven Rostedt wrote:
>
> > On Wed, 14 Feb 2024 12:11:53 +
> > Jonathan Cameron wrote:
> >
> > > So I'm thinking this is a won't fix - wait for the printk rework to land
> > > and
> > > assume this will be resolved as well?
Ira Weiny wrote:
> Jonathan Cameron wrote:
> > On Tue, 06 Feb 2024 14:15:32 -0800
> > Ira Weiny wrote:
> >
> >
[snip]
> >
> > diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c
> > index 9ff8a439d674..7ee45f22f56f 100644
> > --- a/
Jonathan Cameron wrote:
> On Tue, 06 Feb 2024 14:15:32 -0800
> Ira Weiny wrote:
>
> > Smatch caught that cxl_cper_post_event() is called with a spinlock held
> > or preemption disabled.[1] The callback takes the device lock to
> > perform address translatio
Hi Linus, please pull from:
git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm.git
tags/libnvdimm-for-6.8
... to get updates to the nvdimm tree. They are a mix of bug fixes and updates
to interfaces used by nvdimm.
Updates to interfaces include:
Use the new scope based
Randy Dunlap wrote:
>
>
> On 12/21/23 14:32, Ira Weiny wrote:
> > Randy Dunlap wrote:
> >
> > [snip]
> >
> >> @@ -1656,8 +1664,10 @@ static int select_pmem_id(struct nd_regi
> >> /**
> >> * create_namespace_pmem - validate interle
Ira Weiny wrote:
> Randy Dunlap wrote:
[snip]
> > diff -- a/drivers/nvdimm/dimm_devs.c b/drivers/nvdimm/dimm_devs.c
> > --- a/drivers/nvdimm/dimm_devs.c
> > +++ b/drivers/nvdimm/dimm_devs.c
> > @@ -53,7 +53,10 @@ static int validate_dimm(struct nvdimm_d
> >
&g
t; + * @nd_mapping: container of dpa-resource-root + labels
> * @nd_label: target pmem namespace label to evaluate
> + *
> + * Returns: the created device on success or -errno on error
NIT: should this be ERR_PTR(-errno) on error?
Generally good to me though.
Reviewed-by: Ira Weiny
>
found for return value of
> 'nvdimm_allocated_dpa'
>
> Signed-off-by: Randy Dunlap
> Cc: Dan Williams
> Cc: Vishal Verma
> Cc: Dave Jiang
> Cc: Ira Weiny
> Cc: nvd...@lists.linux.dev
> ---
> drivers/nvdimm/dimm_devs.c | 14 --
> 1 file changed, 12 inse
gt; Signed-off-by: Randy Dunlap
> Cc: Vishal Verma
> Cc: Dan Williams
> Cc: Dave Jiang
> Cc: Ira Weiny
Reviewed-by: Ira Weiny
> Cc: nvd...@lists.linux.dev
> ---
> drivers/nvdimm/btt.c |2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
>
Rafael J. Wysocki wrote:
> On Wed, Oct 18, 2023 at 6:28 AM Dan Williams wrote:
> >
> > Michal Wilczynski wrote:
> > > The new cleanup.h facilities that arrived in v6.5-rc1 can replace the
> > > the usage of devm semantics in acpi_nfit_init_interleave_set(). That
> > > routine appears to only be
dinghao.liu@ wrote:
> > Ira Weiny wrote:
> > > Dinghao Liu wrote:
> >
> > [snip]
> >
> > -static int btt_freelist_init(struct arena_info *arena)
> > +static int btt_freelist_init(struct device *dev, struct arena_info *arena)
> >
> >
Ira Weiny wrote:
> Dinghao Liu wrote:
[snip]
> >
> > -static int btt_maplocks_init(struct arena_info *arena)
> > +static int btt_maplocks_init(struct device *dev, struct arena_info *arena)
> > {
> > u32 i;
> >
> > - arena->map_locks =
Dinghao Liu wrote:
> Resources allocated by kcalloc() in btt_freelist_init(),
> btt_rtt_init(), and btt_maplocks_init() are not correctly
> released in their callers when an error happens. For
> example, when an error happens in btt_freelist_init(), its
> caller discover_arenas() will directly
prevent new ones from appearing.
>
> Link:
> http://lore.kernel.org/r/657897453dda8_269bd29...@dwillia2-mobl3.amr.corp.intel.com.notmuch
> Link:
> http://lore.kernel.org/r/6577b0c2a02df_a04c529...@dwillia2-xfh.jf.intel.com.notmuch
> Cc: Vishal Verma
> Cc: Ira Weiny
Reviewed-b
e, so change these
> where applicable.
>
> Cc: Joao Martins
> Suggested-by: Dan Williams
> Signed-off-by: Vishal Verma
>
Reviewed-by: Ira Weiny
dinghao.liu@ wrote:
> > Dave Jiang wrote:
[snip]
> > That said, this patch does not completely fix freelist from leaking in the
> > following error path.
> >
> > discover_arenas()
> > btt_freelist_init() -> ok (memory allocated)
> > btt_rtt_init() -> fail
> >
Dave Jiang wrote:
>
[snip]
First off thanks for the patch. This code seems to have a few things to
clean up.
>
> On 12/6/23 20:43, Dinghao Liu wrote:
> > When an error happens in btt_freelist_init(), its caller
> > discover_arenas() will directly free arena, which makes
> > arena->freelist
Hi Linus, please pull from:
git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm.git
tags/libnvdimm-for-6.7
... to get updates to the nvdimm tree. They are a mix of bug fixes and updates
to interfaces used by nvdimm.
Bug fixes include:
Fix a sleep during spinlock in
Dan Williams
> Cc: Vishal Verma
> Cc: Dave Jiang
> Cc: Ira Weiny
Tested-by: Ira Weiny
Reviewed-by: Ira Weiny
> Signed-off-by: Greg Kroah-Hartman
> ---
> tools/testing/nvdimm/test/ndtest.c | 17 +
> tools/testing/nvdimm/test/nfit.c | 14 +++---
&
> Fixes: 5212e11fde4d ("nd_btt: atomic sector updates")
> Signed-off-by: Tomas Glozar
Thanks for the clarification.
Reviewed-by: Ira Weiny
Tomas Glozar wrote:
> čt 14. 9. 2023 v 22:18 odesílatel Ira Weiny napsal:
> > Is the bug in 1 of 2 places?
> >
> > 1) When btt_write_pg()->lock_map() (when the number of lanes is < number
> >of cpus) and the lane is acquired is called?
> >
> >
Chen Ni wrote:
> Use devm_kstrdup() instead of kstrdup() and check its return value to
> avoid memory leak.
>
> Fixes: 49bddc73d15c ("libnvdimm/of_pmem: Provide a unique name for bus
> provider")
> Signed-off-by: Chen Ni
Reviewed-by: Ira Weiny
> ---
> Cha
Yu Liao wrote:
> Make use of the struct_size() helper instead of an open-coded version,
> in order to avoid any potential type mistakes or integer overflows that,
> in the worst scenario, could lead to heap overflows.
>
> Signed-off-by: Yu Liao
> Reviewed-by: Dave Jiang
Revi
ddDVKsqQX@kreacher/
>
> Fixes: 2a5ab99847bd ("ACPICA: struct acpi_nfit_interleave: Replace 1-element
> array with flexible array")
> Cc: sta...@vger.kernel.org # v6.4+
> Signed-off-by: Yu Liao
> Reviewed-by: Dave Jiang
Reviewed-by: Ira Weiny
> ---
> v1 ->
Tomáš Glozar wrote:
> From: Tomas Glozar
>
> nd_region_acquire_lane uses get_cpu, which disables preemption. This is
> an issue on PREEMPT_RT kernels, since btt_write_pg and also
> nd_region_acquire_lane itself take a spin lock, resulting in BUG:
> sleeping function called from invalid context.
Dave Jiang wrote:
>
>
> On 9/14/23 00:03, Chen Ni wrote:
[snip]
> > diff --git a/drivers/nvdimm/of_pmem.c b/drivers/nvdimm/of_pmem.c
> > index 1b9f5b8a6167..5765674b36f2 100644
> > --- a/drivers/nvdimm/of_pmem.c
> > +++ b/drivers/nvdimm/of_pmem.c
> > @@ -30,7 +30,13 @@ static int
Chen Ni wrote:
> Check the return value of kstrdup() and add kfree() for kstrdup() to
> avoid memory leak.
>
> Fixes: 49bddc73d15c ("libnvdimm/of_pmem: Provide a unique name for bus
> provider")
> Signed-off-by: Chen Ni
> ---
> Changelog:
>
> v1 -> v2:
>
> 1.Add a fixes tag.
> 2.Update
;)
> Signed-off-by: Jiasheng Jiang
LTGM
Reviewed-by: Ira Weiny
> ---
> Changelog:
>
> v2 -> v3:
>
> 1. Correct the usage of devm_kstrdup().
>
> v1 -> v2:
>
> 1. Replace kstrdup() with devm_kstrdup().
> ---
> drivers/nvdimm/of_pmem.c | 7 ++-
&g
Jiasheng Jiang wrote:
> On Wed, Jun 21, 2023 at 00:04:36 +0800, Ira Weiny wrote:
> > Ira Weiny wrote:
> >> Jiasheng Jiang wrote:
[snip]
> >>
> >> Nice catch!
> >>
> >> However, this free needs to happen in of_pmem_region_remove() as wel
Jiasheng Jiang wrote:
> Replace kstrdup() with devm_kstrdup() to avoid memory leak and
> add check for the return value of the devm_kstrdup() to avoid
> NULL pointer dereference
>
> Fixes: 49bddc73d15c ("libnvdimm/of_pmem: Provide a unique name for bus
> provider")
> Signed-off-by: Jiasheng
Ira Weiny wrote:
> Jiasheng Jiang wrote:
> > Add check for the return value of kstrdup() and return the error
> > if it fails in order to avoid NULL pointer dereference.
> > Moreover, use kfree() in the later error handling in order to avoid
> > memory leak.
&
Jiasheng Jiang wrote:
> Add check for the return value of kstrdup() and return the error
> if it fails in order to avoid NULL pointer dereference.
> Moreover, use kfree() in the later error handling in order to avoid
> memory leak.
>
> Fixes: 49bddc73d15c ("libnvdimm/of_pmem: Provide a unique
Ben Dooks wrote:
> The run_dax() prototype is defined in "bus.h" but drivers/dax/super.c
> does not include this. Include bus.h to silece the following sparse
> warning:
>
> drivers/dax/super.c:337:6: warning: symbol 'run_dax' was not declared. Should
> it be static?
A different version of this
Ben Dooks wrote:
> The nd_class is not used outside of drivers/nvdimm/bus.c and thus sparse
> is generating the following warning. Remove this by making it static:
>
> drivers/nvdimm/bus.c:28:14: warning: symbol 'nd_class' was not declared.
> Should it be static?
Reviewe
later.
Cc: Yongqiang Liu
Cc: Dan Williams
Cc: Paul Cassella
Signed-off-by: Ira Weiny
---
drivers/dax/bus.c | 13 ++---
drivers/dax/cxl.c | 4
drivers/dax/hmem/hmem.c | 3 ---
drivers/dax/pmem.c | 8 +---
4 files changed, 7 insertions(+), 21 deletions(-)
diff
: 09d09e04d2fc ("cxl/dax: Create dax devices for CXL RAM regions")
Cc: Dan Williams
Cc: linux-...@vger.kernel.org
Signed-off-by: Ira Weiny
---
This work was inspired by Yongqiang Liu here:
https://lore.kernel.org/all/20221203095858.612027-1-liuyongqian...@huawei.com/
---
drivers/dax/cxl.c |
")
Cc: Dan Williams
Signed-off-by: Ira Weiny
---
drivers/dax/bus.c | 7 ++-
1 file changed, 6 insertions(+), 1 deletion(-)
diff --git a/drivers/dax/bus.c b/drivers/dax/bus.c
index 227800053309..899e29d107b4 100644
--- a/drivers/dax/bus.c
+++ b/drivers/dax/bus.c
@@ -583,6 +583,7 @@ s
evice to
hmem_register_device")
Reviewed-by: Ira Weiny
Signed-off-by: Yongqiang Liu
---
drivers/dax/hmem/hmem.c | 4 +---
1 file changed, 1 insertion(+), 3 deletions(-)
diff --git a/drivers/dax/hmem/hmem.c b/drivers/dax/hmem/hmem.c
index e5fe8b39fb94..b4831a3d3934 100644
--- a/drivers/dax/h
/20221203095858.612027-1-liuyongqian...@huawei.com/
Signed-off-by: Ira Weiny
---
Ira Weiny (3):
dax/bus: Fix leaked reference in alloc_dax_region()
dax/cxl: Fix refcount leak in cxl_dax_region_probe()
dax/bus: Remove unnecessary reference in alloc_dax_region()
Yongqiang Liu (1):
dax/hmem
Paul Cassella wrote:
> On Fri, 2 Jun 2023, Ira Weiny wrote:
> > Paul Cassella wrote:
> > > On Sat, 3 Dec 2022, Ira Weiny wrote:
> > > > On Sat, Dec 03, 2022 at 09:58:58AM +, Yongqiang Liu wrote:
>
> > > > > We should always ca
Paul Cassella wrote:
> On Sat, 3 Dec 2022, Ira Weiny wrote:
> > On Sat, Dec 03, 2022 at 09:58:58AM +, Yongqiang Liu wrote:
>
> > > We should always call dax_region_put() whenever devm_create_dev_dax()
> > > succeed or fail to avoid refcount leak of dax_regio
ter 'nvdimm'
> description in 'nvdimm_allocated_dpa
>
> Reported-by: Abaci Robot
Link: https://bugzilla.openanolis.cn/show_bug.cgi?id=4118
Reviewed-by: Ira Weiny
> Signed-off-by: Yang Li
> ---
> drivers/nvdimm/dimm_devs.c | 4 ++--
> 1 file changed, 2 insertions(+),
On Sat, Dec 03, 2022 at 09:58:58AM +, Yongqiang Liu wrote:
> We should always call dax_region_put() whenever devm_create_dev_dax()
> succeed or fail to avoid refcount leak of dax_region. Move the return
> value check after dax_region_put().
I think dax_region_put is called from
On Sun, Sep 25, 2022 at 09:26:35PM -0400, Bo Liu wrote:
> Use ida_alloc_xxx()/ida_free() instead of
> ida_simple_get()/ida_simple_remove().
> The latter is deprecated and more verbose.
>
> Signed-off-by: Bo Liu
Reviewed-by: Ira Weiny
> ---
> drivers/dax/super.c | 6 +++
On Fri, Aug 19, 2022 at 10:10:24AM -0700, Davidlohr Bueso wrote:
> With CXL security features, global CPU cache flushing nvdimm requirements
> are no longer specific to that subsystem, even beyond the scope of
> security_ops. CXL will need such semantics for features not necessarily
> limited to
On Fri, Aug 05, 2022 at 01:33:19AM -0400, Bo Liu wrote:
> It's possible that dev_set_name() returns -ENOMEM, catch and handle this.
Did this cause a bug or some other problem when the name was not set?
I think it is an ok change but without digging into the code I'm not clear why
you did this.
mappings can be
> acquired from any context (including interrupts).
>
> Therefore, use kmap_local_page() in copy_string_kernel() instead of
> kmap_atomic().
>
> Tested with xfstests on a QEMU + KVM 32-bits VM booting a kernel with
> HIGHMEM64GB enabled.
>
> Suggested-by:
he mappings can be acquired from any
> context (including interrupts).
>
> Therefore, use kmap_local_page() in exec.c because these mappings are per
> thread, CPU local, and not globally visible.
>
> Tested with xfstests on a QEMU + KVM 32-bits VM booting a kernel with
> HIGHMEM64GB e
On Sun, Apr 24, 2022 at 06:26:55AM +, cgel@gmail.com wrote:
> From: ran jianping
>
> All work currently pending will be done first by calling destroy_workqueue,
> so there is no need to flush it explicitly.
>
> Reported-by: Zeal Robot
> Signed-off-by: ran jianpi
On Thu, Apr 21, 2022 at 08:33:51AM -0700, Dan Williams wrote:
> Lockdep reports the following deadlock scenarios for CXL root device
> power-management, device_prepare(), operations, and device_shutdown()
> operations for 'nd_region' devices:
>
> ---
> Chain exists of:
>_region_key -->
Link:
> https://lore.kernel.org/r/ylf0dewci8myl...@hirez.programming.kicks-ass.net [1]
> Cc: Greg Kroah-Hartman
> Cc: "Rafael J. Wysocki"
> Suggested-by: Peter Zijlstra
> Signed-off-by: Dan Williams
Reviewed-by: Ira Weiny
> ---
> drivers/base/core.c|3 ---
>
On Thu, Apr 21, 2022 at 08:33:39AM -0700, Dan Williams wrote:
> Now that all NVDIMM subsystem locking is validated with custom lock
> classes, there is no need for the custom usage of the lockdep_mutex.
>
> Cc: Vishal Verma
> Cc: Dave Jiang
> Cc: Ira Weiny
> Signed
> dropped.
>
> Cc: Vishal Verma
> Cc: Dave Jiang
> Cc: Ira Weiny
> Signed-off-by: Dan Williams
Reviewed-by: Ira Weiny
> ---
> drivers/acpi/nfit/core.c | 30 +++---
> drivers/acpi/nfit/nfit.h | 24
> 2 files
ink:
> https://lore.kernel.org/r/164982968798.684294.15817853329823976469.st...@dwillia2-desk3.amr.corp.intel.com
> [1]
> Link:
> https://lore.kernel.org/r/ylf0dewci8myl...@hirez.programming.kicks-ass.net [2]
> Suggested-by: Peter Zijlstra
> Cc: Vishal Verma
> Cc: Dave Jiang
> Cc: Ira Wein
On Thu, Apr 21, 2022 at 08:33:23AM -0700, Dan Williams wrote:
> Now that all CXL subsystem locking is validated with custom lock
> classes, there is no need for the custom usage of the lockdep_mutex.
>
> Cc: Alison Schofield
> Cc: Vishal Verma
> Cc: Ira Weiny
> Cc: Ben Wid
ere in
the acpi code. So given that it looks to me like ACPI is just using the
default no validate class...
Reviewed-by: Ira Weiny
> However, the lockdep API does support changing the lock class "live" for
> situations like this. Add a device_lock_set_class() helper that a driver
&g
.kernel.org/r/ylf0dewci8myl...@hirez.programming.kicks-ass.net [2]
> Suggested-by: Peter Zijlstra
> Cc: Ingo Molnar
> Cc: Will Deacon
> Cc: Waiman Long
> Cc: Boqun Feng
> Cc: Alison Schofield
> Cc: Vishal Verma
> Cc: Ira Weiny
> Cc: Ben Widawsky
> Cc: Jonathan Ca
On Wed, Mar 09, 2022 at 09:08:36PM -0800, Dan Williams wrote:
> On Fri, Mar 4, 2022 at 12:38 PM wrote:
> >
> > From: Ira Weiny
> >
> > The function run_dax() was missing a prototype when compiling with
> > warnings.
> >
> > Add bus.h to
From: Ira Weiny
struct dax_device has a member named ops which was undocumented.
Add the kdoc.
Signed-off-by: Ira Weiny
---
drivers/dax/super.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/drivers/dax/super.c b/drivers/dax/super.c
index 5c003cc73d04..2fd3a01ba34b 100644
--- a/drivers
From: Ira Weiny
The function run_dax() was missing a prototype when compiling with
warnings.
Add bus.h to fix this.
Signed-off-by: Ira Weiny
---
drivers/dax/super.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/drivers/dax/super.c b/drivers/dax/super.c
index e3029389d809..5c003cc73d04
On Mon, Feb 14, 2022 at 12:09:54PM -0800, Dan Williams wrote:
> On Mon, Feb 14, 2022 at 9:59 AM Ira Weiny wrote:
> >
> > On Fri, Feb 11, 2022 at 11:11:11PM -0800, Tong Zhang wrote:
> > > A bug can be triggered by following command
> > >
> > >
On Fri, Feb 11, 2022 at 11:11:11PM -0800, Tong Zhang wrote:
> A bug can be triggered by following command
>
> $ modprobe nd_pmem && modprobe -r nd_pmem
>
> [ 10.060014] BUG dax_cache (Not tainted): Objects remaining in dax_cache on
> __kmem_cache_shutdown()
> [ 10.060938] Slab
ng: Function parameter or member
> 'nd_region' not described in 'btt_init'
> drivers/nvdimm/btt.c:1584: warning: Excess function parameter 'maxlane'
> description in 'btt_init'
>
> Reported-by: Abaci Robot
> Signed-off-by: Yang Li
Reviewed-by: Ira Weiny
> ---
> drivers/nvdi
On Thu, Nov 25, 2021 at 03:25:09PM +0100, Thomas Gleixner wrote:
> On Tue, Aug 03 2021 at 21:32, ira weiny wrote:
> > @@ -200,16 +200,14 @@ __setup("init_pkru=", setup_init_pkru);
> > */
> > u32 update_pkey_val(u32 pk_reg, int pkey, unsigned int flags)
>
On Mon, Dec 06, 2021 at 05:54:23PM -0800, 'Ira Weiny' wrote:
[snip]
> >
> > Though, if you look at the xen_pv_evtchn_do_upcall() part where you
> > added this extra invocation you might figure out that adding
> > pkrs_restore_irq() to irqentry_exit_cond_r
Thomas,
Thanks for the review. Sorry for being so late to respond I was sick all last
week and so it took me longer to figure out some of this stuff.
On Thu, Nov 25, 2021 at 03:12:47PM +0100, Thomas Gleixner wrote:
> Ira,
>
> On Tue, Aug 03 2021 at 21:32, ira we
On Tue, Aug 03, 2021 at 09:32:21PM -0700, 'Ira Weiny' wrote:
> From: Ira Weiny
>
> The PKRS MSR is not managed by XSAVE. It is preserved through a context
> switch but this support leaves exception handling code open to memory
> accesses during exceptions.
>
> 2 possible
On Sun, Nov 07, 2021 at 06:20:14PM +0100, Christophe JAILLET wrote:
> Le 07/11/2021 à 18:11, Ira Weiny a écrit :
> > On Sat, Nov 06, 2021 at 06:27:11PM +0100, Christophe JAILLET wrote:
> > > If 'devm_init_badblocks()' fails, a previous 'blk_alloc_disk()' call must
> > >
On Sat, Nov 06, 2021 at 06:27:11PM +0100, Christophe JAILLET wrote:
> If 'devm_init_badblocks()' fails, a previous 'blk_alloc_disk()' call must
> be undone.
I think this is a problem...
>
> Signed-off-by: Christophe JAILLET
> ---
> This patch is speculative. Several fixes on error handling
On Mon, Apr 19, 2021 at 07:27:25PM +0800, Wan Jiabing wrote:
> struct device is declared at 133rd line.
> The declaration here is unnecessary. Remove it.
>
> Signed-off-by: Wan Jiabing
> ---
> include/linux/libnvdimm.h | 1 -
> 1 file changed, 1 deletion(-)
>
> diff --git
On Thu, Apr 01, 2021 at 03:58:23PM -0700, 'Ira Weiny' wrote:
> From: Ira Weiny
>
> Introduce a new page protection mechanism for supervisor pages, Protection Key
> Supervisor (PKS).
Is there any feedback on this series?
Perhaps I should ping for specific feedback or an ack? Maybe
On Thu, Apr 08, 2021 at 06:58:26PM -0700, wangyingji...@126.com wrote:
> From: Yingjie Wang
>
> In nd_dax_probe(), 'nd_dax' is allocated by nd_dax_alloc().
> nd_dax_alloc() may fail and return NULL, so we should better check
Avoid the use of 'we'.
> it's return value to avoid a NULL pointer
From: Ira Weiny
Protection Keys for Supervisor pages (PKS) enables fast, hardware thread
specific, manipulation of permission restrictions on supervisor page
mappings. It uses the same mechanism of Protection Keys as those on
User mappings but applies that mechanism to supervisor mappings using
From: Ira Weiny
Previously if a Protection key fault occurred it indicated something
very wrong because user page mappings are not supposed to be in the
kernel address space.
Now PKey faults may happen on kernel mappings if the feature is enabled.
Remove the warning in the fault path and allow
From: Ira Weiny
The core PKS functionality provides an interface for kernel users to
reserve keys to their domains set up the page tables with those keys and
control access to those domains when needed.
Define test code which exercises the core functionality of PKS via a
debugfs entry. Basic
From: Ira Weiny
The PKRS MSR is not managed by XSAVE. It is preserved through a context
switch but this support leaves exception handling code open to memory
accesses during exceptions.
2 possible places for preserving this state were considered,
irqentry_state_t or pt_regs.[1] pt_regs
From: Ira Weiny
The PKRS MSR is defined as a per-logical-processor register. This
isolates memory access by logical CPU. Unfortunately, the MSR is not
managed by XSAVE. Therefore, tasks must save/restore the MSR value on
context switch.
Define a saved PKRS value in the task struct, as well
-developed-by: Ira Weiny
Signed-off-by: Ira Weiny
Signed-off-by: Fenghua Yu
---
Changes from V5:
Dave Hansen
Remove 'we' from comments
Changes from V4:
From Sean Christopherson
Add what happens if the pkey is violated
Changes from V3:
From Dan
From: Ira Weiny
Protection Keys for Supervisor pages (PKS) enables fast, hardware thread
specific, manipulation of permission restrictions on supervisor page
mappings. It uses the same mechanism of Protection Keys as those on
User mappings but applies that mechanism to supervisor mappings using
From: Ira Weiny
Avoid open coding shift and mask operations by defining and using helper
macros for PKey operations.
Reviewed-by: Dan Williams
Signed-off-by: Ira Weiny
---
Changes from V3:
new patch suggested by Dan Williams to use macros better.
---
arch/x86/include/asm/pgtable.h
From: Ira Weiny
Define a helper, update_pkey_val(), which will be used to support both
Protection Key User (PKU) and the new Protection Key for Supervisor
(PKS) in subsequent patches.
Reviewed-by: Dan Williams
Co-developed-by: Peter Zijlstra
Signed-off-by: Peter Zijlstra
Signed-off-by: Ira
From: Ira Weiny
Introduce a new page protection mechanism for supervisor pages, Protection Key
Supervisor (PKS).
Generally PKS enables protections on 'domains' of supervisor pages to limit
supervisor mode access to pages beyond the normal paging protections. PKS
works in a similar fashion
From: Ira Weiny
Protection Keys User (PKU) and Protection Keys Supervisor (PKS) work in
similar fashions and can share common defines. Specifically PKS and PKU
each have:
1. A single control register
2. The same number of keys
3. The same number of bits in the register
On Wed, Mar 31, 2021 at 12:31:56PM -0700, Dave Hansen wrote:
> On 3/31/21 12:14 PM, ira.we...@intel.com wrote:
> > + * To protect against exceptions having access to this memory we save the
> > + * current running value and sets the PKRS value to be used during the
> > + * exception.
>
> This
From: Ira Weiny
The core PKS functionality provides an interface for kernel users to
reserve keys to their domains set up the page tables with those keys and
control access to those domains when needed.
Define test code which exercises the core functionality of PKS via a
debugfs entry. Basic
From: Ira Weiny
The PKRS MSR is not managed by XSAVE. It is preserved through a context
switch but this support leaves exception handling code open to memory
accesses during exceptions.
2 possible places for preserving this state were considered,
irqentry_state_t or pt_regs.[1] pt_regs
From: Ira Weiny
The PKRS MSR is defined as a per-logical-processor register. This
isolates memory access by logical CPU. Unfortunately, the MSR is not
managed by XSAVE. Therefore, tasks must save/restore the MSR value on
context switch.
Define a saved PKRS value in the task struct, as well
From: Ira Weiny
Protection Keys for Supervisor pages (PKS) enables fast, hardware thread
specific, manipulation of permission restrictions on supervisor page
mappings. It uses the same mechanism of Protection Keys as those on
User mappings but applies that mechanism to supervisor mappings using
-developed-by: Ira Weiny
Signed-off-by: Ira Weiny
Signed-off-by: Fenghua Yu
---
Changes from V4:
From Sean Christopherson
Add what happens if the pkey is violated
Changes from V3:
From Dan Williams
Remove flags from pks_key_alloc
From: Ira Weiny
Define a helper, update_pkey_val(), which will be used to support both
Protection Key User (PKU) and the new Protection Key for Supervisor
(PKS) in subsequent patches.
Reviewed-by: Dan Williams
Co-developed-by: Peter Zijlstra
Signed-off-by: Peter Zijlstra
Signed-off-by: Ira
From: Ira Weiny
Previously if a Protection key fault occurred it indicated something
very wrong because user page mappings are not supposed to be in the
kernel address space.
Now PKey faults may happen on kernel mappings if the feature is enabled.
Remove the warning in the fault path and allow
From: Ira Weiny
Introduce a new page protection mechanism for supervisor pages, Protection Key
Supervisor (PKS).
Generally PKS enables protections on 'domains' of supervisor pages to limit
supervisor mode access to pages beyond the normal paging protections. PKS
works in a similar fashion
From: Ira Weiny
Avoid open coding shift and mask operations by defining and using helper
macros for PKey operations.
Reviewed-by: Dan Williams
Signed-off-by: Ira Weiny
---
Changes from V3:
new patch suggested by Dan Williams to use macros better.
---
arch/x86/include/asm/pgtable.h
From: Ira Weiny
Protection Keys User (PKU) and Protection Keys Supervisor (PKS) work in
similar fashions and can share common defines. Specifically PKS and PKU
each have:
1. A single control register
2. The same number of keys
3. The same number of bits in the register
From: Ira Weiny
Protection Keys for Supervisor pages (PKS) enables fast, hardware thread
specific, manipulation of permission restrictions on supervisor page
mappings. It uses the same mechanism of Protection Keys as those on
User mappings but applies that mechanism to supervisor mappings using
[snip]
The test bot reported build errors on i386 yesterday. Not sure why they were
not caught before...
Anyway that caused me to look at this patch again and I've found a couple of
issues noted below. Combined with Sean's review I'll be re-spinning a new v5.
> diff --git
1 - 100 of 963 matches
Mail list logo
