On Wed, Mar 5, 2025 at 6:59 PM Menglong Dong wrote:
>
> I'm not sure if it works. However, indirect call is also used
> in function graph, so we still have better performance. Isn't it?
>
> Let me have a look at the code of the function graph first :/
Menglong,
Function graph infra isn't going t
On Wed, Mar 5, 2025 at 11:02 PM Steven Rostedt wrote:
>
> On Wed, 5 Mar 2025 09:19:09 +0800
> Menglong Dong wrote:
>
> > Ok, let me explain it from the beginning. (My English is not good,
> > but I'll try to describe it as clear as possible :/)
>
> I always appreciate those who struggle with Engl
On 05.03.25 20:58, Andrii Nakryiko wrote:
On Wed, Mar 5, 2025 at 11:47 AM David Hildenbrand wrote:
On 05.03.25 20:43, Andrii Nakryiko wrote:
On Wed, Mar 5, 2025 at 7:22 AM Oleg Nesterov wrote:
On 03/04, David Hildenbrand wrote:
Currently, uprobe_write_opcode() implements COW-breaking man
On Wed, Mar 5, 2025 at 11:47 AM David Hildenbrand wrote:
>
> On 05.03.25 20:43, Andrii Nakryiko wrote:
> > On Wed, Mar 5, 2025 at 7:22 AM Oleg Nesterov wrote:
> >>
> >> On 03/04, David Hildenbrand wrote:
> >>>
> >>> Currently, uprobe_write_opcode() implements COW-breaking manually, which
> >>> i
On 05.03.25 20:43, Andrii Nakryiko wrote:
On Wed, Mar 5, 2025 at 7:22 AM Oleg Nesterov wrote:
On 03/04, David Hildenbrand wrote:
Currently, uprobe_write_opcode() implements COW-breaking manually, which is
really far from ideal.
To say at least ;)
David, thanks for doing this. I'll try to
On Wed, Mar 5, 2025 at 7:22 AM Oleg Nesterov wrote:
>
> On 03/04, David Hildenbrand wrote:
> >
> > Currently, uprobe_write_opcode() implements COW-breaking manually, which is
> > really far from ideal.
>
> To say at least ;)
>
> David, thanks for doing this. I'll try to read 3/3 tomorrow, but I do
On 05.03.25 20:30, Matthew Wilcox wrote:
On Tue, Mar 04, 2025 at 04:48:46PM +0100, David Hildenbrand wrote:
+static bool orig_page_is_identical(struct vm_area_struct *vma,
+ unsigned long vaddr, struct page *page, bool *pmd_mappable)
+{
+ const pgoff_t index = vaddr_to_offset
The .rodata.(cst|str)* sections are often resized during the final
linking and since these sections do not cover actual symbols there is
no need to include them in the modules.builtin.ranges data.
When these sections were included in processing and resizing occurred,
modules were reported with ran
On Tue, Mar 04, 2025 at 04:48:46PM +0100, David Hildenbrand wrote:
> +static bool orig_page_is_identical(struct vm_area_struct *vma,
> + unsigned long vaddr, struct page *page, bool *pmd_mappable)
> +{
> + const pgoff_t index = vaddr_to_offset(vma, vaddr) >> PAGE_SHIFT;
> + stru
Add 2 per-cpu monitors as part of the sched model:
* sco: scheduling context operations
Monitor to ensure sched_set_state happens only in thread context
* tss: task switch while scheduling
Monitor to ensure sched_switch happens only in scheduling context
To: Ingo Molnar
To: Peter Zijlstr
From: Steven Rostedt
There's no reason to save the KASLR offset for the ring buffer itself.
That is used by the tracer. Now that the tracer has a way to save data in
the persistent memory of the ring buffer, have the tracing infrastructure
take care of the saving of the KASLR offset.
Reviewed-by
From: Steven Rostedt
Add the last boot module's names and addresses to the last_boot_info file.
This only shows the module information from a previous boot. If the buffer
is started and is recording the current boot, this file still will only
show "current".
~# cat instances/boot_mapped/last_b
This updates the persistent instance to record what modules were
loaded and what addresses they were loaded at.
First the KASLR offset is recorded in the persistent ring buffer instead of
a text address. This can be used to calculated the address offset.
Next the persistent memory is divided up
From: Steven Rostedt
Instead of just having a meta data at the first page of each sub buffer
that has duplicate data, add a new meta page to the entire block of memory
that holds the duplicate data and remove it from the sub buffer meta data.
This will open up the extra memory in this first page
From: Steven Rostedt
The tracing system needs a way to save all the currently loaded modules
and their addresses into persistent memory so that it can evaluate the
addresses on a reboot from a crash. When the persistent memory trace
starts, it will load the module addresses and names into the per
From: Steven Rostedt
When a module is loaded and a persistent buffer is actively tracing, add
it to the list of modules in the persistent memory.
Signed-off-by: Steven Rostedt (Google)
---
kernel/trace/trace.c| 27 +
kernel/trace/trace.h| 2 ++
kernel/t
From: Steven Rostedt
Now that there's one meta data at the start of the persistent memory used by
the ring buffer, allow the caller to request some memory right after that
data that it can use as its own persistent memory.
Also fix some white space issues with ring_buffer_alloc().
Reviewed-by:
From: Steven Rostedt
Instead of saving off the text and data pointers and using them to compare
with the current boot's text and data pointers, just save off the KASLR
offset. Then that can be used to figure out how to read the previous boots
buffer.
The last_boot_info will now show this offset,
From: Steven Rostedt
For trace instances that are mapped to persistent memory, have them use
the scratch area to save the currently loaded modules. This will allow
where the modules have been loaded on the next boot so that their
addresses can be deciphered by using where they were loaded previou
Add man page and kernel documentation for the sched monitors, as sched
is a container of other monitors, document all in the same page.
sched is the first nested monitor, also explain what is a nested monitor
and how enabling containers or children monitors work.
To: Ingo Molnar
To: Peter Zijlstr
On 03/04, David Hildenbrand wrote:
>
> Currently, uprobe_write_opcode() implements COW-breaking manually, which is
> really far from ideal.
To say at least ;)
David, thanks for doing this. I'll try to read 3/3 tomorrow, but I don't
think I can really help. Let me repeat, this code was written man
On Wed, 5 Mar 2025 08:53:43 -0500
Mathieu Desnoyers wrote:
> On 2025-03-04 20:55, Huang Shijie wrote:
> > When tracepoint_debug is set, we may get the output in kernel log:
> > [ 380.013843] Probe 0 : f0d68cda
> >
> > It is not readable, so change to print the function symbol.
> >
On Wed, 5 Mar 2025 09:47:49 +0100
Petr Pavlu wrote:
> On 3/4/25 15:57, Steven Rostedt wrote:
> >
> > Luis,
> >
> > Can I get an Acked-by from you?
> >
> > This follows the changes you have in linux-next.
>
> Hi Steven,
>
> I'm not Luis, but we started rotating the modules maintenance every
On Wed, 5 Mar 2025 09:19:09 +0800
Menglong Dong wrote:
> Ok, let me explain it from the beginning. (My English is not good,
> but I'll try to describe it as clear as possible :/)
I always appreciate those who struggle with English having these
conversations. Thank you for that, as I know I am ho
Hi.
On Fri, Feb 21, 2025 at 04:18:54PM -0800, Andrew Morton
wrote:
> This seems thin. Is there more we can tell our users? What are the
> visible effects of this? What are the benefits? Why would they want
> to turn it on?
Thanks for review and comments (also Alexander).
> I mean, there ar
On Sat, Feb 22, 2025 at 09:02:08AM +, David Laight
wrote:
> It also seems a good way of being able to predict the next pid and
> doing all the 'nasty' things that allows because there is no guard
> time on pid reuse.
The motivations was not to make guessing next pid more difficult, I'll
upda
RV now supports nested monitors, this functionality requires a container
monitor, which has virtually no functionality besides holding other
monitors, and nested monitors, that have a container as parent.
Add the -p flag to pass a parent to a monitor, this sets it up while
registering the monitor
Add 3 per-cpu monitors as part of the sched model:
* scpd: schedule called with preemption disabled
Monitor to ensure schedule is called with preemption disabled
* snep: schedule does not enable preempt
Monitor to ensure schedule does not enable preempt
* sncid: schedule not called with in
Monitors describing complex systems, such as the scheduler, can easily
grow to the point where they are just hard to understand because of the
many possible state transitions.
Often it is possible to break such descriptions into smaller monitors,
sharing some or all events. Enabling those smaller m
This patchset starts including adapted scheduler specifications from
Daniel's task model [1].
As the model is fairly complicated, it is split in several generators
and specifications. The tool used to create the model can output a
unified model, but that would be hardly readable (9k states).
RV al
Add the following tracepoints:
* sched_entry(bool preempt, ip)
Called while entering __schedule
* sched_exit(bool is_switch, ip)
Called while exiting __schedule
* sched_set_state(task, curr_state, state)
Called when a task changes its state (to and from running)
These tracepoints are u
Add possibility to supply the container name to rv list:
# rv list sched
mon1
mon2
mon3
This lists only monitors in sched, without indentation.
Supplying -h, any option (string starting with -) or more than 1
argument will still print the usage.
Passing a non-existent container prints not
RV now supports nested monitors, this functionality requires a container
monitor, which has virtually no functionality besides holding other
monitors, and nested monitors, that have a container as parent.
Nested monitors' sysfs folders are physically nested in the container's
folder, and they are
Add a per-task monitor as part of the sched model:
* snroc: set non runnable on its own context
Monitor to ensure set_state happens only in the respective task's context
To: Ingo Molnar
To: Peter Zijlstra
Cc: Juri Lelli
Cc: John Kacur
Cc: Clark Williams
Signed-off-by: Gabriele Monaco
--
On 2025-03-04 20:55, Huang Shijie wrote:
When tracepoint_debug is set, we may get the output in kernel log:
[ 380.013843] Probe 0 : f0d68cda
It is not readable, so change to print the function symbol.
After this patch, the output may becomes:
[ 54.930567] Probe 0 : perf_tr
On Sat, Mar 01, 2025 at 09:23:51AM +0200, Mike Rapoport wrote:
> Hi Brendan,
>
> On Fri, Jan 10, 2025 at 06:40:28PM +, Brendan Jackman wrote:
> > Currently a nop config. Keeping as a separate commit for easy review of
> > the boring bits. Later commits will use and enable this new config.
> >
>
On Tue, 2025-03-04 at 12:18 -0500, Steven Rostedt wrote:
> On Tue, 18 Feb 2025 13:31:07 +0100
> Gabriele Monaco wrote:
>
> > Some monitor files like the main header and the Kconfig are missing
> > the
> > license identifier.
> >
> > Add it to those and make sure the automatic generation script
>
On 04/03/25 12:43 am, Nico Pache wrote:
On Wed, Feb 19, 2025 at 9:52 AM Ryan Roberts wrote:
On 11/02/2025 00:30, Nico Pache wrote:
Introduce the ability for khugepaged to collapse to different mTHP sizes.
While scanning a PMD range for potential collapse candidates, keep track
of pages in
On 19/02/25 10:22 pm, Ryan Roberts wrote:
On 11/02/2025 00:30, Nico Pache wrote:
Introduce the ability for khugepaged to collapse to different mTHP sizes.
While scanning a PMD range for potential collapse candidates, keep track
of pages in MIN_MTHP_ORDER chunks via a bitmap. Each bit represen
On Wed, Mar 5, 2025 at 4:30 PM H. Peter Anvin wrote:
>
> On March 4, 2025 5:19:09 PM PST, Menglong Dong
> wrote:
> >On Tue, Mar 4, 2025 at 10:53 PM H. Peter Anvin wrote:
> >>
> >> On March 4, 2025 1:42:20 AM PST, Peter Zijlstra
> >> wrote:
> >> >On Tue, Mar 04, 2025 at 03:47:45PM +0800, Mengl
On 3/4/25 15:57, Steven Rostedt wrote:
>
> Luis,
>
> Can I get an Acked-by from you?
>
> This follows the changes you have in linux-next.
Hi Steven,
I'm not Luis, but we started rotating the modules maintenance every six
months and I'm currently looking after the modules tree. I see Luis
seeme
On March 4, 2025 5:19:09 PM PST, Menglong Dong wrote:
>On Tue, Mar 4, 2025 at 10:53 PM H. Peter Anvin wrote:
>>
>> On March 4, 2025 1:42:20 AM PST, Peter Zijlstra wrote:
>> >On Tue, Mar 04, 2025 at 03:47:45PM +0800, Menglong Dong wrote:
>> >> We don't have to select FUNCTION_ALIGNMENT_32B, so th
42 matches
Mail list logo
