On Mon, Oct 28, 2024 at 08:44:10AM -1000, Linus Torvalds wrote:
> I The *hardware* presumably uses either bit 47/56 because that's the
> actual hardware width of the TLB / cache matching.
It is a guess, right?
The bug makes more sense to me if the bit number depends on the active
paging mode: thi
On Mon, 28 Oct 2024 at 10:31, Kirill A. Shutemov wrote:
>
> On Mon, Oct 28, 2024 at 08:44:10AM -1000, Linus Torvalds wrote:
> > I The *hardware* presumably uses either bit 47/56 because that's the
> > actual hardware width of the TLB / cache matching.
>
> It is a guess, right?
Yes, I have no idea
Implement the ASRC memory to memory function using
the compress framework, user can use this function with
compress ioctl interface.
This feature can be shared by ASRC and EASRC drivers
Signed-off-by: Shengjiu Wang
Acked-by: Jaroslav Kysela
---
sound/soc/fsl/Kconfig | 1 +
sound/so
ASRC can be used on memory to memory case, define several
functions for m2m usage.
m2m_prepare: prepare for the start step
m2m_start: the start step
m2m_unprepare: unprepare for stop step, optional
m2m_stop: stop step
m2m_check_format: check format is supported or not
m2m_calc_out_len: calculate o
This function is base on the accelerator implementation
for compress API:
https://patchwork.kernel.org/project/alsa-devel/patch/20241002093904.1809799-1-pe...@perex.cz/
Audio signal processing also has the requirement for memory to
memory similar as Video.
This asrc memory to memory (memory ->asr
ASRC can be used on memory to memory case, define several
functions for m2m usage and export them as function pointer.
Signed-off-by: Shengjiu Wang
Acked-by: Jaroslav Kysela
---
sound/soc/fsl/fsl_easrc.c | 228 ++
sound/soc/fsl/fsl_easrc.h | 4 +
2 files ch
Register m2m platform device, that user can
use M2M feature.
Signed-off-by: Shengjiu Wang
Acked-by: Jaroslav Kysela
---
sound/soc/fsl/fsl_asrc.c | 37 +
1 file changed, 33 insertions(+), 4 deletions(-)
diff --git a/sound/soc/fsl/fsl_asrc.c b/sound/soc/fsl/fs
Register m2m platform device,that user can
use M2M feature.
Signed-off-by: Shengjiu Wang
Acked-by: Jaroslav Kysela
---
sound/soc/fsl/fsl_easrc.c | 33 +++--
1 file changed, 31 insertions(+), 2 deletions(-)
diff --git a/sound/soc/fsl/fsl_easrc.c b/sound/soc/fsl/fsl_e
Add 'pcm_format' for struct snd_codec, add 'pcm_formats' for
struct snd_codec_desc, these are used for accelerator usage.
Current accelerator example is sample rate converter (SRC).
Define struct snd_codec_desc_src for descript minmum and maxmum
sample rates. And add 'src_d' in union snd_codec_opt
Alistair Popple wrote:
[..]
> >> > It follows that that the DMA-idle condition still needs to look for the
> >> > case where the refcount is > 1 rather than 0 since refcount == 1 is the
> >> > page-mapped-but-DMA-idle condition.
>
> Because if the DAX page-cache holds a reference the refcount won
The barrier_nospec() in 64-bit copy_from_user() is slow. Instead use
pointer masking to force the user pointer to all 1's if the access_ok()
mispredicted true for an invalid address.
The kernel test robot reports a 2.6% improvement in the per_thread_ops
benchmark (see link below).
To avoid regre
v3:
- rebase on Linus' patch (v6.12-rc5)
- remove putuser(), copy_to_user(), clear_user() changes: as Linus
points out, they don't produce a value that could be used in a
dependent load to leak anything.
- convert 32-bit versions of copy_from_user(), get_user() and
__get_user() and converg
On Mon, Oct 28, 2024 at 06:56:15PM -0700, Josh Poimboeuf wrote:
> The barrier_nospec() in 64-bit __get_user() is slow. Instead use
> pointer masking to force the user pointer to all 1's if a previous
> access_ok() mispredicted true for an invalid address.
Linus pointed out that __get_user() may b
The x86 implementations of get_user() and __get_user() are now
identical. Merge their implementations and make the __get_user()
implementations aliases of their get_user() counterparts.
Signed-off-by: Josh Poimboeuf
---
arch/x86/lib/getuser.S | 58 +-
1 f
The barrier_nospec() in 34-bit __get_user() is slow. Instead use
pointer masking to force the user pointer to all 1's if the access_ok()
mispredicted true for an invalid address.
Signed-off-by: Josh Poimboeuf
---
arch/x86/lib/getuser.S | 18 --
1 file changed, 18 deletions(-)
d
The barrier_nospec() in 32-bit copy_from_user() is slow. Instead use
pointer masking to force the user pointer to all 1's if a previous
access_ok() mispredicted true for an invalid address.
Signed-off-by: Josh Poimboeuf
---
arch/x86/include/asm/uaccess.h| 34 +++
The barrier_nospec() in 64-bit __get_user() is slow. Instead use
pointer masking to force the user pointer to all 1's if a previous
access_ok() mispredicted true for an invalid address.
Note that for safety on some AMD CPUs, this relies on recent commit
86e6b1547b3d ("x86: fix user address maskin
On architectures where ARCH_HAS_SYNC_CORE_BEFORE_USERMODE
is not selected, sync_core_before_usermode() is a no-op.
In membarrier_mm_sync_core_before_usermode() the compiler does not
eliminate redundant branches and load of mm->membarrier_state
for this case as the atomic_read() cannot be optimized
On 28/10/24 7:53 am, Michael Ellerman wrote:
Hari Bathini writes:
From: Naveen N Rao
Add support for bpf_arch_text_poke() and arch_prepare_bpf_trampoline()
for 64-bit powerpc. While the code is generic, BPF trampolines are only
enabled on 64-bit powerpc. 32-bit powerpc will need testing an
Convert the 32-bit get_user() implementations to use the new
unconditional masking scheme for consistency with 64-bit.
Signed-off-by: Josh Poimboeuf
---
arch/x86/lib/getuser.S | 33 ++---
1 file changed, 14 insertions(+), 19 deletions(-)
diff --git a/arch/x86/lib/get
The Power11 architected and raw mode support in Linux was merged via [1]
and the corresponding support in QEMU is pending in [2], which is
currently in its V6.
Currently, booting a KVM guest inside a pseries LPAR (Logical Partition)
on a kernel without [1] results the guest boot in a Power10
compa
Running a L2 vCPU (see [1] for terminology) with LPCR_MER bit set and no
pending interrupts results in that L2 vCPU getting an infinite flood of
spurious interrupts. The 'if check' in kvmhv_run_single_vcpu() sets the
LPCR_MER bit if there are pending interrupts.
The spurious flood problem can be o
On 28/10/2024 10:07, Yury Khrustalev wrote:
> Replace literal 0 with macro PKEY_UNRESTRICTED where pkey_*() functions
> are used in mm selftests for memory protection keys for ppc target.
>
> Signed-off-by: Yury Khrustalev
> Suggested-by: Kevin Brodsky
Reviewed-by: Kevin Brodsky
- Kevin
From: Shengjiu Wang
[ Upstream commit 54c805c1eb264c839fa3027d0073bb7f323b0722 ]
Irq handler need to be executed as fast as possible, so
the log in irq handler is better to use dev_dbg which needs
to be enabled when debugging.
Signed-off-by: Shengjiu Wang
Reviewed-by: Iuliana Prodan
Link:
ht
From: Shengjiu Wang
[ Upstream commit b9a8ecf81066e01e8a3de35517481bc5aa0439e5 ]
On some platforms, for example i.MX93, there is only one
audio PLL source, so some sample rate can't be supported.
If the PLL source is used for 8kHz series rates, then 11kHz
series rates can't be supported.
So add
From: Shengjiu Wang
[ Upstream commit 54c805c1eb264c839fa3027d0073bb7f323b0722 ]
Irq handler need to be executed as fast as possible, so
the log in irq handler is better to use dev_dbg which needs
to be enabled when debugging.
Signed-off-by: Shengjiu Wang
Reviewed-by: Iuliana Prodan
Link:
ht
From: Shengjiu Wang
[ Upstream commit 54c805c1eb264c839fa3027d0073bb7f323b0722 ]
Irq handler need to be executed as fast as possible, so
the log in irq handler is better to use dev_dbg which needs
to be enabled when debugging.
Signed-off-by: Shengjiu Wang
Reviewed-by: Iuliana Prodan
Link:
ht
From: Shengjiu Wang
[ Upstream commit 54c805c1eb264c839fa3027d0073bb7f323b0722 ]
Irq handler need to be executed as fast as possible, so
the log in irq handler is better to use dev_dbg which needs
to be enabled when debugging.
Signed-off-by: Shengjiu Wang
Reviewed-by: Iuliana Prodan
Link:
ht
From: Shengjiu Wang
[ Upstream commit 54c805c1eb264c839fa3027d0073bb7f323b0722 ]
Irq handler need to be executed as fast as possible, so
the log in irq handler is better to use dev_dbg which needs
to be enabled when debugging.
Signed-off-by: Shengjiu Wang
Reviewed-by: Iuliana Prodan
Link:
ht
On 25/10/2024 23:15, Rob Herring wrote:
> On Wed, Oct 23, 2024 at 12:14 PM Usama Arif wrote:
>>
>> __pa() is only intended to be used for linear map addresses and using
>> it for initial_boot_params which is in fixmap for arm64 will give an
>> incorrect value. Hence save the physical address w
On Mon, Oct 28, 2024 at 10:41 AM Will Deacon wrote:
>
> On Wed, Oct 23, 2024 at 06:14:26PM +0100, Usama Arif wrote:
> > __pa() is only intended to be used for linear map addresses and using
> > it for initial_boot_params which is in fixmap for arm64 will give an
> > incorrect value. Hence save th
On Wed, Oct 23, 2024 at 06:14:26PM +0100, Usama Arif wrote:
> __pa() is only intended to be used for linear map addresses and using
> it for initial_boot_params which is in fixmap for arm64 will give an
> incorrect value. Hence save the physical address when it is known at
> boot time when calling
On Thu, Oct 10, 2024 at 09:01:07AM +0200, Thomas Weißschuh wrote:
> The vvar mapping is the same for all processes. Use a single mapping to
> simplify the logic and align it with the other architectures.
>
> In addition this will enable the move of the vvar handling into generic code.
>
> Signed-
On Mon, 28 Oct 2024 at 01:29, Kirill A. Shutemov wrote:
>
> I think it is worth looking at this approach again as it gets better
> code: removes two instructions from get_user() and doesn't get flags
> involved.
The problem with the 'and' is that it's just re-introducing the same old bug.
> The
On Thu, Oct 10, 2024 at 09:01:06AM +0200, Thomas Weißschuh wrote:
> This constant is always "0", providing no value and making the logic
> harder to understand.
> Also prepare for a consolidation of the vdso linkerscript logic by
> aligning it with other architectures.
>
> Signed-off-by: Thomas We
From: Shengjiu Wang
[ Upstream commit 54c805c1eb264c839fa3027d0073bb7f323b0722 ]
Irq handler need to be executed as fast as possible, so
the log in irq handler is better to use dev_dbg which needs
to be enabled when debugging.
Signed-off-by: Shengjiu Wang
Reviewed-by: Iuliana Prodan
Link:
ht
From: Shengjiu Wang
[ Upstream commit 54c805c1eb264c839fa3027d0073bb7f323b0722 ]
Irq handler need to be executed as fast as possible, so
the log in irq handler is better to use dev_dbg which needs
to be enabled when debugging.
Signed-off-by: Shengjiu Wang
Reviewed-by: Iuliana Prodan
Link:
ht
On Mon, Oct 28, 2024 at 06:49:47AM -0400, Sasha Levin wrote:
> From: Shengjiu Wang
>
> [ Upstream commit 54c805c1eb264c839fa3027d0073bb7f323b0722 ]
>
> Irq handler need to be executed as fast as possible, so
> the log in irq handler is better to use dev_dbg which needs
> to be enabled when debug
On Wed, Oct 16, 2024 at 03:34:11PM -0700, Linus Torvalds wrote:
> On Wed, 16 Oct 2024 at 15:13, Kirill A. Shutemov wrote:
> >
> > It is worse than that. If we get LAM_SUP enabled (there's KASAN patchset
> > in works) this check will allow arbitrary kernel addresses.
>
> Ugh. I haven't seen the LA
From: Shengjiu Wang
[ Upstream commit b9a8ecf81066e01e8a3de35517481bc5aa0439e5 ]
On some platforms, for example i.MX93, there is only one
audio PLL source, so some sample rate can't be supported.
If the PLL source is used for 8kHz series rates, then 11kHz
series rates can't be supported.
So add
Replace literal 0 with macro PKEY_UNRESTRICTED where pkey_*() functions
are used in mm selftests for memory protection keys.
Signed-off-by: Yury Khrustalev
Suggested-by: Joey Gouly
---
tools/testing/selftests/mm/mseal_test.c| 6 +++---
tools/testing/selftests/mm/pkey-helpers.h
Memory protection keys (pkeys) uapi has two macros for pkeys restrictions:
- PKEY_DISABLE_ACCESS 0x1
- PKEY_DISABLE_WRITE 0x2
with implicit literal value of 0x0 that means "unrestricted". Code that
works with pkeys has to use this literal value when implying that a pkey
imposes no restrictions
Replace literal 0 with macro PKEY_UNRESTRICTED where pkey_*() functions
are used in mm selftests for memory protection keys for ppc target.
Signed-off-by: Yury Khrustalev
Suggested-by: Kevin Brodsky
---
Note that I couldn't build these tests so I would appreciate if someone
could check this pat
Add PKEY_UNRESTRICTED macro to mman.h and use it in selftests.
For context, this change will also allow for more consistent update of the
Glibc manual [1] which in turn will help with introducing memory protection
keys on AArch64 targets [2].
Applies to 42f7652d3eb5 (tag: v6.12-rc4).
Note that I
44 matches
Mail list logo
