On 2024/8/2 11:45, Song Liu wrote:
>
>
>> On Aug 1, 2024, at 6:18 PM, Leizhen (ThunderTown)
>> wrote:
>>
>> On 2024/7/31 9:00, Song Liu wrote:
>>> Hi Masami,
>>>
On Jul 30, 2024, at 6:03 AM, Masami Hiramatsu wrote:
On Mon, 29 Jul 2024 17:54:32 -0700
Song Liu wrote:
>>>
On Fri, 2024-08-02 at 11:29 +0800, Jason Wang wrote:
> On Thu, Aug 1, 2024 at 11:38 PM Dragos Tatulea wrote:
> >
> > The following workflow triggers the crash referenced below:
> >
> > 1) vhost_vdpa_unsetup_vq_irq() unregisters the irq bypass producer
> >but the producer->token is still vali
> Subject: Re: [PATCH v2 2/2] remoteproc: imx_rproc: handle system off
> for i.MX7ULP
>
> On Tue, Jul 30, 2024 at 08:06:22AM +, Peng Fan wrote:
> > > Subject: Re: [PATCH v2 2/2] remoteproc: imx_rproc: handle system
> off
> > > for i.MX7ULP
> > >
> > > On Fri, Jul 19, 2024 at 04:49:04PM +0800,
> On Aug 1, 2024, at 6:18 PM, Leizhen (ThunderTown)
> wrote:
>
> On 2024/7/31 9:00, Song Liu wrote:
>> Hi Masami,
>>
>>> On Jul 30, 2024, at 6:03 AM, Masami Hiramatsu wrote:
>>>
>>> On Mon, 29 Jul 2024 17:54:32 -0700
>>> Song Liu wrote:
>>>
With CONFIG_LTO_CLANG=y, the compiler may
On Thu, Aug 1, 2024 at 11:38 PM Dragos Tatulea wrote:
>
> The following workflow triggers the crash referenced below:
>
> 1) vhost_vdpa_unsetup_vq_irq() unregisters the irq bypass producer
>but the producer->token is still valid.
> 2) vq context gets released and reassigned to another vq.
Jus
On Thu, Aug 1, 2024 at 3:00 PM Michael S. Tsirkin wrote:
>
> On Thu, Aug 01, 2024 at 02:55:10PM +0800, Jason Wang wrote:
> > On Thu, Aug 1, 2024 at 2:42 PM Michael S. Tsirkin wrote:
> > >
> > > On Thu, Aug 01, 2024 at 02:13:18PM +0800, Jason Wang wrote:
> > > > On Thu, Aug 1, 2024 at 1:58 PM Mich
在 2024/8/1 5:42, Andrii Nakryiko 写道:
> From: Peter Zijlstra
>
> With uprobe_unregister() having grown a synchronize_srcu(), it becomes
> fairly slow to call. Esp. since both users of this API call it in a
> loop.
>
> Peel off the sync_srcu() and do it once, after the loop.
>
> With recent up
在 2024/8/1 22:06, Oleg Nesterov 写道:
> On 08/01, Liao Chang wrote:
>>
>> @@ -2276,22 +2277,25 @@ static void handle_singlestep(struct uprobe_task
>> *utask, struct pt_regs *regs)
>> int err = 0;
>>
>> uprobe = utask->active_uprobe;
>> -if (utask->state == UTASK_SSTEP_ACK)
>> +
在 2024/8/2 0:49, Andrii Nakryiko 写道:
> On Thu, Aug 1, 2024 at 5:23 AM Liao, Chang wrote:
>>
>>
>>
>> 在 2024/8/1 5:42, Andrii Nakryiko 写道:
>>> To avoid unnecessarily taking a (brief) refcount on uprobe during
>>> breakpoint handling in handle_swbp for entry uprobes, make find_uprobe()
>>> not ta
Hi Chenyu,
On 2024/8/1 下午10:40, Chen Yu wrote:
Hi Bibo,
On 2024-08-01 at 16:00:19 +0800, maobibo wrote:
Chenyu,
I do not know much about x86, just give some comments(probably incorrected)
from the code.
On 2024/7/29 下午2:52, Chen Yu wrote:
X86_FEATURE_HYPERVISOR YYY N
CON
On 2024/7/31 9:00, Song Liu wrote:
> Hi Masami,
>
>> On Jul 30, 2024, at 6:03 AM, Masami Hiramatsu wrote:
>>
>> On Mon, 29 Jul 2024 17:54:32 -0700
>> Song Liu wrote:
>>
>>> With CONFIG_LTO_CLANG=y, the compiler may add suffix to function names
>>> to avoid duplication. This causes confusion
On Wed, Jul 31, 2024 at 2:43 PM Andrii Nakryiko wrote:
>
> Revamp how struct uprobe is refcounted, and thus how its lifetime is
> managed.
>
> Right now, there are a few possible "owners" of uprobe refcount:
> - uprobes_tree RB tree assumes one refcount when uprobe is registered
> and added
On Thu, Aug 1, 2024 at 11:58 AM Andrii Nakryiko
wrote:
>
> + bpf
>
> On Thu, Aug 1, 2024 at 6:36 AM Peter Zijlstra wrote:
> >
> > On Thu, Aug 01, 2024 at 03:26:38PM +0200, Oleg Nesterov wrote:
> > > (Andrii, I'll try to look at your new series on Weekend).
> >
> > OK, I dropped all your previous
Hi Petr,
On Thu, Aug 1, 2024 at 4:22 AM Petr Pavlu wrote:
>
> STG is an interesting tool. I've played with it a bit last year. To be
> frank, I was surprised to see a new tool being proposed by Google to
> generate modversion CRCs from DWARF instead of potentially extending
> your STG project for
On 01.08.24 17:11, Andreas Hindborg wrote:
> "Benno Lossin" writes:
>> On 01.08.24 15:40, Andreas Hindborg wrote:
>>> "Benno Lossin" writes:
On 01.08.24 13:29, Andreas Hindborg wrote:
> "Benno Lossin" writes:
>> On 05.07.24 13:15, Andreas Hindborg wrote:
>>> +
>>> +/// Types
+ bpf
On Thu, Aug 1, 2024 at 6:36 AM Peter Zijlstra wrote:
>
> On Thu, Aug 01, 2024 at 03:26:38PM +0200, Oleg Nesterov wrote:
> > (Andrii, I'll try to look at your new series on Weekend).
>
> OK, I dropped all your previous patches and stuffed these in.
>
> They should all be visible in queue/per
On Thu, Aug 1, 2024 at 12:28 PM Peter Zijlstra wrote:
>
> On Wed, Jul 31, 2024 at 11:34:13PM +0200, Alice Ryhl wrote:
>
> > > Please work harder to not have to duplicate stuff like this.
> >
> > I really didn't want to duplicate it, but it's very hard to find a
> > performant alternative. Is there
On Thu, Aug 01, 2024 at 11:35:05AM +0200, Peter Zijlstra wrote:
> On Wed, Jul 31, 2024 at 02:42:56PM -0700, Andrii Nakryiko wrote:
> > This patch switches uprobes SRCU usage to RCU Tasks Trace flavor, which
> > is optimized for more lightweight and quick readers (at the expense of
> > slower writer
Hello RT-list!
I'm pleased to announce the 5.10.222-rt114 stable release.
This release is just an update to the new stable 5.10.222 version and
no RT specific changes have been made.
You can get this release via the git tree at:
git://git.kernel.org/pub/scm/linux/kernel/git/rt/linux-stable-rt
On Thu, Aug 1, 2024 at 7:27 AM Jiri Olsa wrote:
>
> On Wed, Jul 31, 2024 at 02:42:53PM -0700, Andrii Nakryiko wrote:
>
> SNIP
>
> > static int __copy_insn(struct address_space *mapping, struct file *filp,
> > void *insn, int nbytes, loff_t offset)
> > {
> > @@ -924,7 +901,8
On Thu, Aug 1, 2024 at 5:23 AM Liao, Chang wrote:
>
>
>
> 在 2024/8/1 5:42, Andrii Nakryiko 写道:
> > To avoid unnecessarily taking a (brief) refcount on uprobe during
> > breakpoint handling in handle_swbp for entry uprobes, make find_uprobe()
> > not take refcount, but protect the lifetime of a upr
On Thu, Aug 1, 2024 at 4:09 AM Jiri Olsa wrote:
>
> On Wed, Jul 31, 2024 at 02:42:50PM -0700, Andrii Nakryiko wrote:
>
> SNIP
>
> > static void put_uprobe(struct uprobe *uprobe)
> > {
> > - if (refcount_dec_and_test(&uprobe->ref)) {
> > - /*
> > - * If application mu
On Thu, Aug 1, 2024 at 2:35 AM Peter Zijlstra wrote:
>
> On Wed, Jul 31, 2024 at 02:42:56PM -0700, Andrii Nakryiko wrote:
> > This patch switches uprobes SRCU usage to RCU Tasks Trace flavor, which
> > is optimized for more lightweight and quick readers (at the expense of
> > slower writers, which
On Tue, Jul 30, 2024 at 09:43:05PM +0200, Luigi Leonardi via B4 Relay wrote:
> This patch series introduce the support for ioctl(s) in AF_VSOCK.
> The only ioctl currently available is SIOCOUTQ, which returns
> the number of unsent or unacked packets. It is available for
> SOCK_STREAM, SOCK_SEQPACK
The following workflow triggers the crash referenced below:
1) vhost_vdpa_unsetup_vq_irq() unregisters the irq bypass producer
but the producer->token is still valid.
2) vq context gets released and reassigned to another vq.
3) That other vq registers it's producer with the same vq context
p
On Thu, Aug 1, 2024 at 6:01 AM Yunsheng Lin wrote:
>
> On 2024/8/1 2:13, Alexander Duyck wrote:
> > On Wed, Jul 31, 2024 at 5:50 AM Yunsheng Lin wrote:
> >>
> >> Currently the page_frag API is returning 'virtual address'
> >> or 'va' when allocing and expecting 'virtual address' or
> >> 'va' as i
"Benno Lossin" writes:
> On 01.08.24 15:40, Andreas Hindborg wrote:
>> "Benno Lossin" writes:
>>> On 01.08.24 13:29, Andreas Hindborg wrote:
"Benno Lossin" writes:
> On 05.07.24 13:15, Andreas Hindborg wrote:
>> +
>> +/// Types that can be used for module parameters.
>> +//
Hi Bibo,
On 2024-08-01 at 16:00:19 +0800, maobibo wrote:
> Chenyu,
>
> I do not know much about x86, just give some comments(probably incorrected)
> from the code.
>
> On 2024/7/29 下午2:52, Chen Yu wrote:
> > X86_FEATURE_HYPERVISOR YYY N
> > CONFIG_PARAVIRT_SPINLOCKS Y
On Wed, Jul 31, 2024 at 02:42:53PM -0700, Andrii Nakryiko wrote:
SNIP
> static int __copy_insn(struct address_space *mapping, struct file *filp,
> void *insn, int nbytes, loff_t offset)
> {
> @@ -924,7 +901,8 @@ static bool filter_chain(struct uprobe *uprobe, struct
> mm_
On 01.08.24 15:40, Andreas Hindborg wrote:
> "Benno Lossin" writes:
>> On 01.08.24 13:29, Andreas Hindborg wrote:
>>> "Benno Lossin" writes:
On 05.07.24 13:15, Andreas Hindborg wrote:
> +
> +/// Types that can be used for module parameters.
> +///
> +/// Note that displaying
On 08/01, Liao Chang wrote:
>
> @@ -2276,22 +2277,25 @@ static void handle_singlestep(struct uprobe_task
> *utask, struct pt_regs *regs)
> int err = 0;
>
> uprobe = utask->active_uprobe;
> - if (utask->state == UTASK_SSTEP_ACK)
> + switch (utask->state) {
> + case UTASK_S
"Benno Lossin" writes:
> On 01.08.24 13:29, Andreas Hindborg wrote:
>>
>> Hi Benno,
>>
>> Thanks for the comments!
>>
>> "Benno Lossin" writes:
>>
>>> On 05.07.24 13:15, Andreas Hindborg wrote:
>>
>> [...]
>>
+
+/// Types that can be used for module parameters.
+///
+//
On Thu, Aug 01, 2024 at 03:26:38PM +0200, Oleg Nesterov wrote:
> (Andrii, I'll try to look at your new series on Weekend).
OK, I dropped all your previous patches and stuffed these in.
They should all be visible in queue/perf/core, and provided the robot
doesn't scream, I'll push them into tip/pe
Kill the extra get_uprobe() + put_uprobe() in uprobe_unregister() and
move the possibly final put_uprobe() from delete_uprobe() to its only
caller, uprobe_unregister().
Signed-off-by: Oleg Nesterov
Acked-by: Andrii Nakryiko
Acked-by: Masami Hiramatsu (Google)
Reviewed-by: Jiri Olsa
---
kernel
Fold __uprobe_unregister() into its single caller, uprobe_unregister().
A separate patch to simplify the next change.
Signed-off-by: Oleg Nesterov
Acked-by: Andrii Nakryiko
Acked-by: Masami Hiramatsu (Google)
Reviewed-by: Jiri Olsa
---
kernel/events/uprobes.c | 25 ++---
1
If register_for_each_vma() fails uprobe_register() can safely drop
uprobe->register_rwsem and use uprobe_unregister(). There is no worry
about the races with another register/unregister, consumer_add() was
already called so this case doesn't differ from _unregister() right
after the successful _reg
This way uprobe_unregister() and uprobe_apply() can use "struct uprobe *"
rather than inode + offset. This simplifies the code and allows to avoid
the unnecessary find_uprobe() + put_uprobe() in these functions.
TODO: uprobe_unregister() still needs get_uprobe/put_uprobe to ensure that
this uprobe
It doesn't make any sense to have 2 versions of _register(). Note that
trace_uprobe_enable(), the only user of uprobe_register(), doesn't need
to check tu->ref_ctr_offset to decide which one should be used, it could
safely pass ref_ctr_offset == 0 to uprobe_register_refctr().
Add this argument to
From: Jiri Olsa
testmod_unregister_uprobe() forgets to path_put(&uprobe.path).
Signed-off-by: Jiri Olsa
Signed-off-by: Oleg Nesterov
---
tools/testing/selftests/bpf/bpf_testmod/bpf_testmod.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/tools/testing/selftests/bpf/bpf_testmod/bpf_testmo
From: Andrii Nakryiko
Return -ENOMEM instead of NULL, which makes caller's error handling just
a touch simpler.
Signed-off-by: Andrii Nakryiko
Signed-off-by: Oleg Nesterov
Reviewed-by: Masami Hiramatsu (Google)
Reviewed-by: Jiri Olsa
---
kernel/events/uprobes.c | 4 +---
1 file changed, 1 i
get_user_pages_remote() and the comment above it make no sense.
There is no task_struct passed into get_user_pages_remote() anymore, and
nowadays mm_account_fault() increments the current->min/maj_flt counters
regardless of FAULT_FLAG_REMOTE.
Reported-by: Andrii Nakryiko
Signed-off-by: Oleg Nest
The comment above uprobe_write_opcode() is wrong, unapply_uprobe() calls
it under mmap_read_lock() and this is correct.
And it is completely unclear why register_for_each_vma() takes mmap_lock
for writing, add a comment to explain that mmap_write_lock() is needed to
avoid the following race:
(Andrii, I'll try to look at your new series on Weekend).
Changes:
- added the acks I got from Andrii, Masami, and Jiri
- new 4/9 patch from Jiri, fixes the unrelated bug in bpf_testmod
- adapt tools/testing/selftests/bpf/bpf_testmod/bpf_testmod.c
to the API chan
On Tue, Jul 30, 2024 at 08:06:22AM +, Peng Fan wrote:
> > Subject: Re: [PATCH v2 2/2] remoteproc: imx_rproc: handle system off
> > for i.MX7ULP
> >
> > On Fri, Jul 19, 2024 at 04:49:04PM +0800, Peng Fan (OSS) wrote:
> > > From: Peng Fan
> > >
> > > The i.MX7ULP Cortex-A7 is under control of C
On Wed, Jul 31, 2024 at 07:17:33PM +0200, Oleg Nesterov wrote:
> On 07/31, Peter Zijlstra wrote:
> >
> > On Wed, Jul 31, 2024 at 10:01:47AM -0700, Andrii Nakryiko wrote:
> > > > Do I stuff this on top of Oleg's patch or do you want me to fold it in
> > > > one of them?
> > >
> > > Please fold so we
On Thu, Aug 01, 2024 at 02:26:45PM +0200, Oleg Nesterov wrote:
> On 08/01, Jiri Olsa wrote:
> >
> > > Note the additional path_put() in testmod_unregister_uprobe(). Does it
> > > need
> > > a separate patch or can it come with 5/5 ?
> >
> > I think it'd be better to have it separately, the test is
On 2024/8/1 2:13, Alexander Duyck wrote:
> On Wed, Jul 31, 2024 at 5:50 AM Yunsheng Lin wrote:
>>
>> Currently the page_frag API is returning 'virtual address'
>> or 'va' when allocing and expecting 'virtual address' or
>> 'va' as input when freeing.
>>
>> As we are about to support new use cases
On 08/01, Jiri Olsa wrote:
>
> > Note the additional path_put() in testmod_unregister_uprobe(). Does it need
> > a separate patch or can it come with 5/5 ?
>
> I think it'd be better to have it separately, the test is already
> released.. so people might want to backport just the fix
OK, I'll reba
On 01.08.24 13:29, Andreas Hindborg wrote:
>
> Hi Benno,
>
> Thanks for the comments!
>
> "Benno Lossin" writes:
>
>> On 05.07.24 13:15, Andreas Hindborg wrote:
>
> [...]
>
>>> +
>>> +/// Types that can be used for module parameters.
>>> +///
>>> +/// Note that displaying the type in `sysfs`
在 2024/8/1 5:42, Andrii Nakryiko 写道:
> To avoid unnecessarily taking a (brief) refcount on uprobe during
> breakpoint handling in handle_swbp for entry uprobes, make find_uprobe()
> not take refcount, but protect the lifetime of a uprobe instance with
> RCU. This improves scalability, as refcoun
On Thu, Aug 01, 2024 at 02:00:18PM +0200, Oleg Nesterov wrote:
> On 08/01, Jiri Olsa wrote:
> >
> > > @@ -474,10 +477,10 @@ static void testmod_unregister_uprobe(void)
> > > {
> > > mutex_lock(&testmod_uprobe_mutex);
> > >
> > > - if (uprobe.offset) {
> > > - uprobe_unr
On 08/01, Jiri Olsa wrote:
>
> > @@ -474,10 +477,10 @@ static void testmod_unregister_uprobe(void)
> > {
> > mutex_lock(&testmod_uprobe_mutex);
> >
> > - if (uprobe.offset) {
> > - uprobe_unregister(d_real_inode(uprobe.path.dentry),
> > -
On Wed, Jul 31, 2024 at 09:18:00AM -0700, Andrii Nakryiko wrote:
SNIP
> diff --git a/tools/testing/selftests/bpf/bpf_testmod/bpf_testmod.c
> b/tools/testing/selftests/bpf/bpf_testmod/bpf_testmod.c
> index 5f152afdec2f..73a6b041bcce 100644
> --- a/tools/testing/selftests/bpf/bpf_testmod/bpf_testmo
Hi Benno,
Thanks for the comments!
"Benno Lossin" writes:
> On 05.07.24 13:15, Andreas Hindborg wrote:
[...]
>> +
>> +/// Types that can be used for module parameters.
>> +///
>> +/// Note that displaying the type in `sysfs` will fail if
>> +/// [`core::str::from_utf8`] (as implemented thro
On 7/26/24 23:05, Sami Tolvanen wrote:
> On Mon, Jul 22, 2024 at 8:20 AM Petr Pavlu wrote:
>>
>> From my perspective, I'm okay if gendwarfksyms doesn't provide
>> functionality to compare a new object file with its reference symtypes
>> file.
>>
>> As mentioned, genksyms has this functionality but
On Wed, Jul 31, 2024 at 02:42:50PM -0700, Andrii Nakryiko wrote:
SNIP
> static void put_uprobe(struct uprobe *uprobe)
> {
> - if (refcount_dec_and_test(&uprobe->ref)) {
> - /*
> - * If application munmap(exec_vma) before uprobe_unregister()
> - * gets c
On Wed, Jul 31, 2024 at 11:34:13PM +0200, Alice Ryhl wrote:
> > Please work harder to not have to duplicate stuff like this.
>
> I really didn't want to duplicate it, but it's very hard to find a
> performant alternative. Is there any way we could accept duplication
> only in the cases where an '
>
> Could probably do:
>
> #ifdef HAVE_LIBCPUPOWER_SUPPORT
> > + "--deepest-idle-state n: only go down to idle
> > state n on cpus used by timerlat to reduce exit from idle latency",
> #else
> + "--deepest-idle-state n: [rtla built without
> libcp
>
> Need to check if the calloc failed and return an error if it did.
>
Definitely, I completely missed that.
>
> Hmm, should this warn if state is not zero and disabled is negative.
>
rtla timerlat hist/top will error out if save_cpu_idle_disable_state
returns a negative value, so I don't think
On Wed, Jul 31, 2024 at 02:42:56PM -0700, Andrii Nakryiko wrote:
> This patch switches uprobes SRCU usage to RCU Tasks Trace flavor, which
> is optimized for more lightweight and quick readers (at the expense of
> slower writers, which for uprobes is a fine tradeof) and has better
> performance and
The profiling result of BPF selftest on ARM64 platform reveals the
significant contention on the current->sighand->siglock within the
handle_singlestep() is the scalability bottleneck. The reason is also
very straightforward that all producer threads of benchmark have to
contend the spinlock mentio
Chenyu,
I do not know much about x86, just give some comments(probably
incorrected) from the code.
On 2024/7/29 下午2:52, Chen Yu wrote:
The kernel can change spinlock behavior when running as a guest. But
this guest-friendly behavior causes performance problems on bare metal.
So there's a 'vir
On Thu, Aug 01, 2024 at 02:55:10PM +0800, Jason Wang wrote:
> On Thu, Aug 1, 2024 at 2:42 PM Michael S. Tsirkin wrote:
> >
> > On Thu, Aug 01, 2024 at 02:13:18PM +0800, Jason Wang wrote:
> > > On Thu, Aug 1, 2024 at 1:58 PM Michael S. Tsirkin wrote:
> > > >
> > > > On Thu, Aug 01, 2024 at 10:16:0
63 matches
Mail list logo
