Skip to contentSkip to conversation

/proc/pid/status: Add support for architecture specific output

PATCH In review Aubrey Li 3 participants 14 messages Feb 12, 2019

Subsystems (3)
lore
Share & tools
lore

Aubrey Li

author
#0 lore raw

PATCH v9 1/3 /proc/pid/status: Add support for architecture specific output

The architecture specific information of the running processes could be useful to the userland. Add support to examine process architecture specific information externally.

Cc Peter Zijlstra <[email protected]>
Cc Andi Kleen <[email protected]>
Cc Tim Chen <[email protected]>
Cc Dave Hansen <[email protected]>
Cc Arjan van de Ven <[email protected]>
---
 fs/proc/array.c | 5 +++++
 1 file changed, 5 insertions(+)
fs/proc/array.c +5 -0
index 0ceb3b6b37e7..d8cb5b5fd7bb 100644
@@ -392,6 +392,10 @@ static inline void task_core_dumping(struct seq_file *m, struct mm_struct *mm)
 	seq_putc(m, '\n');
 }
 
+void __weak arch_proc_pid_status(struct seq_file *m, struct task_struct *task)
+{
+}
+
 int proc_pid_status(struct seq_file *m, struct pid_namespace *ns,
 			struct pid *pid, struct task_struct *task)
 {
@@ -414,6 +418,7 @@ int proc_pid_status(struct seq_file *m, struct pid_namespace *ns,
 	task_cpus_allowed(m, task);
 	cpuset_task_status_allowed(m, task);
 	task_context_switch_counts(m, task);
+	arch_proc_pid_status(m, task);
 	return 0;
 }
 
Signature · 1 line
2.17.1
author
#1 lore raw

PATCH v9 2/3 x86,/proc/pid/status: Add AVX-512 usage elapsed time

AVX-512 components use could cause core turbo frequency drop. So it's useful to expose AVX-512 usage elapsed time as a heuristic hint for the user space job scheduler to cluster the AVX-512 using tasks together.

Example:

$ cat /proc/pid/status | grep AVX512_elapsed_ms
AVX512_elapsed_ms:      1020

The number '1020' denotes 1020 millisecond elapsed since last time context switch the off-CPU task using AVX-512 components, thus the task could cause core frequency drop.

Or:

$ cat /proc/pid/status | grep AVX512_elapsed_ms
AVX512_elapsed_ms:      -1

The number '-1' indicates the task didn't use AVX-512 components before thus unlikely has frequency drop issue.

User space tools may want to further check by:

$ perf stat --pid <pid> -e core_power.lvl2_turbo_license -- sleep 1

Performance counter stats for process id '3558':

     3,251,565,961      core_power.lvl2_turbo_license

       1.004031387 seconds time elapsed

Non-zero counter value confirms that the task causes frequency drop.

Cc Peter Zijlstra <[email protected]>
Cc Andi Kleen <[email protected]>
Cc Tim Chen <[email protected]>
Cc Dave Hansen <[email protected]>
Cc Arjan van de Ven <[email protected]>
---
 arch/x86/include/asm/processor.h |  2 ++
 arch/x86/kernel/fpu/xstate.c     | 41 ++++++++++++++++++++++++++++++++
 2 files changed, 43 insertions(+)
arch/x86/include/asm/processor.h +2 -0
index d53c54b842da..60ee932070fe 100644
@@ -996,5 +996,7 @@ enum l1tf_mitigations {
 };
 
 extern enum l1tf_mitigations l1tf_mitigation;
+/* Add support for architecture specific output in /proc/pid/status */
+extern void arch_proc_pid_status(struct seq_file *m, struct task_struct *task);
 
 #endif /* _ASM_X86_PROCESSOR_H */
arch/x86/kernel/fpu/xstate.c +41 -0

50 lines not on this page · raw message

Signature · 1 line
2.17.1
#2 lore raw

Re: [PATCH v9 2/3] x86,/proc/pid/status: Add AVX-512 usage elapsed time

arch/x86/include/asm/processor.h enum l1tf_mitigations { · line 1000

Quoted 10 lines · Aubrey Li
arch/x86/include/asm/processor.h +2 -0
index d53c54b842da..60ee932070fe 100644
@@ -996,5 +996,7 @@ enum l1tf_mitigations {
 };
 
 extern enum l1tf_mitigations l1tf_mitigation;
+/* Add support for architecture specific output in /proc/pid/status */
+extern void arch_proc_pid_status(struct seq_file *m, struct task_struct *task);

Sigh. This is absolutely the wrong place. The weak function is declared and used in fs/proc/... So the prototype wants to be in a header which is included from there independent of x86...

Thanks,

tglx

#3 lore raw

Re: [PATCH v9 2/3] x86,/proc/pid/status: Add AVX-512 usage elapsed time

Quoted 8 lines · Thomas Gleixner
Quoted 7 lines · Aubrey Li
arch/x86/include/asm/processor.h +0 -0
index d53c54b842da..60ee932070fe 100644
@@ -996,5 +996,7 @@ enum l1tf_mitigations {
 };
 
Quoted 7 lines

arch/x86/include/asm/processor.h enum l1tf_mitigations { · line 1000

Quoted 3 lines
arch/x86/include/asm/processor.h +2 -0
 extern enum l1tf_mitigations l1tf_mitigation;
+/* Add support for architecture specific output in /proc/pid/status */
+extern void arch_proc_pid_status(struct seq_file *m, struct task_struct *task);

Sigh. This is absolutely the wrong place. The weak function is declared and used in fs/proc/... So the prototype wants to be in a header which is included from there independent of x86...

Can the prototype be in the architecture header if they want to call the function? Like the following? arch_report_meminfo() is used in fs/proc/... as well.

$ find . -name *.h | xargs grep arch_report_meminfo
./arch/s390/include/asm/pgtable.h:void arch_report_meminfo(struct seq_file *m);
./arch/x86/include/asm/pgtable_types.h:extern void arch_report_meminfo(struct seq_file *m);
./arch/parisc/include/asm/pgtable.h:extern void arch_report_meminfo(struct seq_file *m);

Thanks,
-Aubrey

#4 lore raw

Re: [PATCH v9 2/3] x86,/proc/pid/status: Add AVX-512 usage elapsed time

Quoted 17 lines · Aubrey
Quoted 8 lines · Thomas Gleixner
Quoted 7 lines · Aubrey Li
arch/x86/include/asm/processor.h +0 -0
index d53c54b842da..60ee932070fe 100644
@@ -996,5 +996,7 @@ enum l1tf_mitigations {
 };
 
Quoted 7 lines

arch/x86/include/asm/processor.h enum l1tf_mitigations { · line 1000

Quoted 3 lines
arch/x86/include/asm/processor.h +2 -0
 extern enum l1tf_mitigations l1tf_mitigation;
+/* Add support for architecture specific output in /proc/pid/status */
+extern void arch_proc_pid_status(struct seq_file *m, struct task_struct *task);

Sigh. This is absolutely the wrong place. The weak function is declared and used in fs/proc/... So the prototype wants to be in a header which is included from there independent of x86...

Quoted 8 lines

Can the prototype be in the architecture header if they want to call the function? Like the following? arch_report_meminfo() is used in fs/proc/... as well.

$ find . -name *.h | xargs grep arch_report_meminfo
./arch/s390/include/asm/pgtable.h:void arch_report_meminfo(struct seq_file *m);
./arch/x86/include/asm/pgtable_types.h:extern void arch_report_meminfo(struct seq_file *m);
./arch/parisc/include/asm/pgtable.h:extern void arch_report_meminfo(struct seq_file *m);

Actually both way exist in the current kernel,the reason I chose to put the prototype into architecture header file is that I found some architectures rename the function name by a micro definition while others use prototype. See below:

$ find . -name *.h | xargs grep arch_irq_stat
./arch/arm64/include/asm/hardirq.h:#define arch_irq_stat_cpu	smp_irq_stat_cpu
./arch/arm/include/asm/hardirq.h:#define arch_irq_stat_cpu	smp_irq_stat_cpu
./arch/x86/include/asm/hardirq.h:extern u64 arch_irq_stat_cpu(unsigned int cpu);

This looks more flexible than it in the common header file.

Anyway, putting the prototype into the common header file like include/linux/proc_fs.h is also acceptable to me if you persist, please just let me know, :)

Thanks,
-Aubrey

#5 lore raw

Re: [PATCH v9 2/3] x86,/proc/pid/status: Add AVX-512 usage elapsed time

Quoted 6 lines · Aubrey
$ find . -name *.h | xargs grep arch_irq_stat
./arch/arm64/include/asm/hardirq.h:#define arch_irq_stat_cpu	smp_irq_stat_cpu
./arch/arm/include/asm/hardirq.h:#define arch_irq_stat_cpu	smp_irq_stat_cpu
./arch/x86/include/asm/hardirq.h:extern u64 arch_irq_stat_cpu(unsigned int cpu);

This looks more flexible than it in the common header file.

Looking more flexible does not make it more correct.

Thanks,

tglx

#6 lore raw

Re: [PATCH v9 2/3] x86,/proc/pid/status: Add AVX-512 usage elapsed time

Quoted 9 lines · Thomas Gleixner
Quoted 6 lines · Aubrey
$ find . -name *.h | xargs grep arch_irq_stat
./arch/arm64/include/asm/hardirq.h:#define arch_irq_stat_cpu	smp_irq_stat_cpu
./arch/arm/include/asm/hardirq.h:#define arch_irq_stat_cpu	smp_irq_stat_cpu
./arch/x86/include/asm/hardirq.h:extern u64 arch_irq_stat_cpu(unsigned int cpu);

This looks more flexible than it in the common header file.

Looking more flexible does not make it more correct.

Okay, will cook a new version to put it into the common header.

Thanks,
-Aubrey

#7 lore raw

Re: [PATCH v9 2/3] x86,/proc/pid/status: Add AVX-512 usage elapsed time

In reply to Li, Aubrey
Quoted 17 lines · Aubrey
Quoted 8 lines · Thomas Gleixner
Quoted 7 lines · Aubrey Li
arch/x86/include/asm/processor.h +0 -0
index d53c54b842da..60ee932070fe 100644
@@ -996,5 +996,7 @@ enum l1tf_mitigations {
 };
 
Quoted 7 lines

arch/x86/include/asm/processor.h enum l1tf_mitigations { · line 1000

Quoted 3 lines
arch/x86/include/asm/processor.h +2 -0
 extern enum l1tf_mitigations l1tf_mitigation;
+/* Add support for architecture specific output in /proc/pid/status */
+extern void arch_proc_pid_status(struct seq_file *m, struct task_struct *task);

Sigh. This is absolutely the wrong place. The weak function is declared and used in fs/proc/... So the prototype wants to be in a header which is included from there independent of x86...

Quoted 2 lines

Can the prototype be in the architecture header if they want to call the function?

Basic C programming course:

The prototype must be available before the declaration of the global function.

fs/proc/array.c:404:13: warning: no previous prototype for ‘arch_proc_pid_status’ [-Wmissing-prototypes]
void __weak arch_proc_pid_status(struct seq_file *m, struct task_struct *task)

Oh well....

Thanks,

tglx

#8 lore raw

Re: [PATCH v9 2/3] x86,/proc/pid/status: Add AVX-512 usage elapsed time

Quoted 23 lines · Thomas Gleixner
Quoted 17 lines · Aubrey
Quoted 8 lines · Thomas Gleixner
Quoted 7 lines · Aubrey Li
arch/x86/include/asm/processor.h +0 -0
index d53c54b842da..60ee932070fe 100644
@@ -996,5 +996,7 @@ enum l1tf_mitigations {
 };
 
Quoted 7 lines

arch/x86/include/asm/processor.h enum l1tf_mitigations { · line 1000

Quoted 3 lines
arch/x86/include/asm/processor.h +2 -0
 extern enum l1tf_mitigations l1tf_mitigation;
+/* Add support for architecture specific output in /proc/pid/status */
+extern void arch_proc_pid_status(struct seq_file *m, struct task_struct *task);

Sigh. This is absolutely the wrong place. The weak function is declared and used in fs/proc/... So the prototype wants to be in a header which is included from there independent of x86...

Quoted 2 lines

Can the prototype be in the architecture header if they want to call the function?

Basic C programming course:

Quoted 7 lines

The prototype must be available before the declaration of the global function.

fs/proc/array.c:404:13: warning: no previous prototype for ‘arch_proc_pid_status’ [-Wmissing-prototypes]
void __weak arch_proc_pid_status(struct seq_file *m, struct task_struct *task)

Oh well....

Is this because patch 1/3 applied alone? If the whole patch set are applied, the prototype is included in <asm/processor.h>, which is at the beginning of array.c file, so it is available before the declaration.

Thanks,
-Aubrey

#9 lore raw

Re: [PATCH v9 2/3] x86,/proc/pid/status: Add AVX-512 usage elapsed time

Quoted 27 lines · Aubrey
Quoted 26 lines · Thomas Gleixner
Quoted 17 lines · Aubrey
Quoted 8 lines · Thomas Gleixner
Quoted 7 lines · Aubrey Li
arch/x86/include/asm/processor.h +0 -0
index d53c54b842da..60ee932070fe 100644
@@ -996,5 +996,7 @@ enum l1tf_mitigations {
 };
 
Quoted 7 lines

arch/x86/include/asm/processor.h enum l1tf_mitigations { · line 1000

Quoted 3 lines
arch/x86/include/asm/processor.h +2 -0
 extern enum l1tf_mitigations l1tf_mitigation;
+/* Add support for architecture specific output in /proc/pid/status */
+extern void arch_proc_pid_status(struct seq_file *m, struct task_struct *task);

Sigh. This is absolutely the wrong place. The weak function is declared and used in fs/proc/... So the prototype wants to be in a header which is included from there independent of x86...

Quoted 2 lines

Can the prototype be in the architecture header if they want to call the function?

Basic C programming course:

The prototype must be available before the declaration of the global function.

Quoted 8 lines
Quoted 4 lines

fs/proc/array.c:404:13: warning: no previous prototype for ‘arch_proc_pid_status’ [-Wmissing-prototypes]
void __weak arch_proc_pid_status(struct seq_file *m, struct task_struct *task)

Oh well....

Is this because patch 1/3 applied alone? If the whole patch set are applied, the prototype is included in <asm/processor.h>, which is at the beginning of array.c file, so it is available before the declaration.

1) Each patch has to be correct stand alone

2) This file is compiled for every architecture the kernel supports and how many of them are including arch/x86/include/asm/processor.h ?

There is a world outside x86 and it's rather large.

Thanks,

tglx

#10 lore raw

Re: [PATCH v9 2/3] x86,/proc/pid/status: Add AVX-512 usage elapsed time

Quoted 37 lines · Thomas Gleixner
Quoted 27 lines · Aubrey
Quoted 26 lines · Thomas Gleixner
Quoted 17 lines · Aubrey
Quoted 8 lines · Thomas Gleixner
Quoted 7 lines · Aubrey Li
arch/x86/include/asm/processor.h +0 -0
index d53c54b842da..60ee932070fe 100644
@@ -996,5 +996,7 @@ enum l1tf_mitigations {
 };
 
Quoted 7 lines

arch/x86/include/asm/processor.h enum l1tf_mitigations { · line 1000

Quoted 3 lines
arch/x86/include/asm/processor.h +2 -0
 extern enum l1tf_mitigations l1tf_mitigation;
+/* Add support for architecture specific output in /proc/pid/status */
+extern void arch_proc_pid_status(struct seq_file *m, struct task_struct *task);

Sigh. This is absolutely the wrong place. The weak function is declared and used in fs/proc/... So the prototype wants to be in a header which is included from there independent of x86...

Quoted 2 lines

Can the prototype be in the architecture header if they want to call the function?

Basic C programming course:

The prototype must be available before the declaration of the global function.

Quoted 8 lines
Quoted 4 lines

fs/proc/array.c:404:13: warning: no previous prototype for ‘arch_proc_pid_status’ [-Wmissing-prototypes]
void __weak arch_proc_pid_status(struct seq_file *m, struct task_struct *task)

Oh well....

Is this because patch 1/3 applied alone? If the whole patch set are applied, the prototype is included in <asm/processor.h>, which is at the beginning of array.c file, so it is available before the declaration.

Quoted 6 lines

1) Each patch has to be correct stand alone

2) This file is compiled for every architecture the kernel supports and how many of them are including arch/x86/include/asm/processor.h ?

There is a world outside x86 and it's rather large.

Got it, thanks!

Quoted 5 lines

Thanks,

tglx

author
#11 lore raw

PATCH v9 3/3 Documentation/filesystems/proc.txt: add AVX512_elapsed_ms

In reply to Aubrey Li

Added AVX512_elapsed_ms in /proc/<pid>/status. Report it in Documentation/filesystems/proc.txt

Cc Peter Zijlstra <[email protected]>
Cc Andi Kleen <[email protected]>
Cc Tim Chen <[email protected]>
Cc Dave Hansen <[email protected]>
Cc Arjan van de Ven <[email protected]>
---
 Documentation/filesystems/proc.txt | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
Documentation/filesystems/proc.txt +3 -1
index 520f6a84cf50..8275ccd766e4 100644
@@ -197,6 +197,7 @@ read the file /proc/PID/status:
   Seccomp:        0
   voluntary_ctxt_switches:        0
   nonvoluntary_ctxt_switches:     1
+  AVX512_elapsed_ms:	1020
 
 This shows you nearly the same information you would get if you viewed it with
 the ps  command.  In  fact,  ps  uses  the  proc  file  system  to  obtain its
@@ -214,7 +215,7 @@ asynchronous manner and the value may not be very precise. To see a precise
 snapshot of a moment, you can see /proc/<pid>/smaps file and scan page table.
 It's slow but very precise.
 
-Table 1-2: Contents of the status files (as of 4.8)
+Table 1-2: Contents of the status files (as of 5.1)
 ..............................................................................
  Field                       Content
  Name                        filename of the executable
@@ -275,6 +276,7 @@ Table 1-2: Contents of the status files (as of 4.8)
  Mems_allowed_list           Same as previous, but in "list format"
  voluntary_ctxt_switches     number of voluntary context switches
  nonvoluntary_ctxt_switches  number of non voluntary context switches
+ AVX512_elapsed_ms           time elapsed since last AVX512 use in millisecond
 ..............................................................................
 
 Table 1-3: Contents of the statm files (as of 2.6.8-rc3)
Signature · 1 line
2.17.1
#12 lore raw
In reply to Aubrey Li
Quoted 19 lines · Aubrey Li

The architecture specific information of the running processes could be useful to the userland. Add support to examine process architecture specific information externally.

Cc Peter Zijlstra <[email protected]>
Cc Andi Kleen <[email protected]>
Cc Tim Chen <[email protected]>
Cc Dave Hansen <[email protected]>
Cc Arjan van de Ven <[email protected]>
---
 fs/proc/array.c | 5 +++++
 1 file changed, 5 insertions(+)
fs/proc/array.c +0 -0
index 0ceb3b6b37e7..d8cb5b5fd7bb 100644

fs/proc/array.c static inline void task_core_dumping(struct seq_file *m, struct mm_struct *mm) · line 397

Quoted 6 lines
fs/proc/array.c +3 -0
@@ -392,6 +392,10 @@ static inline void task_core_dumping(struct seq_file *m, struct mm_struct *mm)
 	seq_putc(m, '\n');
 }
 
+void __weak arch_proc_pid_status(struct seq_file *m, struct task_struct *task)
+{
+}

This still lacks a prototype.

Thanks,

tglx

#13 lore raw
Quoted 24 lines · Thomas Gleixner
Quoted 22 lines · Aubrey Li

The architecture specific information of the running processes could be useful to the userland. Add support to examine process architecture specific information externally.

Cc Peter Zijlstra <[email protected]>
Cc Andi Kleen <[email protected]>
Cc Tim Chen <[email protected]>
Cc Dave Hansen <[email protected]>
Cc Arjan van de Ven <[email protected]>
---
 fs/proc/array.c | 5 +++++
 1 file changed, 5 insertions(+)
fs/proc/array.c +0 -0
index 0ceb3b6b37e7..d8cb5b5fd7bb 100644
@@ -392,6 +392,10 @@ static inline void task_core_dumping(struct seq_file *m, struct mm_struct *mm)
 	seq_putc(m, '\n');
 }
 
Quoted 5 lines

fs/proc/array.c static inline void task_core_dumping(struct seq_file *m, struct mm_struct *mm) · line 397

Quoted 3 lines
fs/proc/array.c +3 -0
+void __weak arch_proc_pid_status(struct seq_file *m, struct task_struct *task)
+{
+}

This still lacks a prototype.

Aside of that this doesn't apply to tip or Linus tree....

Thanks,

tglx