perf test record.sh: Raise limit of open file descriptors
PATCH v2 1/1 perf test record.sh: Raise limit of open file descriptors
From: Veronika Molnarova <[email protected]>
Subtest for system-wide record with '--threads=cpu' option fails due to a limit of open file descriptors on systems with 128 or more CPUs as the default limit is set to 1024.
The number of open file descriptors should be slightly above nmb_events*nmb_cpus + nmb_cpus(for perf.data.n) + 4*nmb_cpus(for pipes), which equals 8*nmb_cpus. Therefore, temporarily raise the limit to 16*nmb_cpus for the test.
--- tools/perf/tests/shell/record.sh | 16 ++++++++++++++++ 1 file changed, 16 insertions(+)
tools/perf/tests/shell/record.sh +16 -0
@@ -11,6 +11,14 @@ err=0 perfdata=$(mktemp /tmp/__perf_test.perf.data.XXXXX) testprog="perf test -w thloop" testsym="test_loop" +default_fd_limit=$(ulimit -n) +# With option --threads=cpu the number of open file descriptors should be +# equal to sum of: nmb_cpus * nmb_events (2+dummy), +# nmb_threads for perf.data.n (equal to nmb_cpus) and +# 2*nmb_cpus of pipes = 4*nmb_cpus (each pipe has 2 ends) +# All together it needs 8*nmb_cpus file descriptors plus some are also used +# outside of testing, thus raising the limit to 16*nmb_cpus +min_fd_limit=$(($(getconf _NPROCESSORS_ONLN) * 16)) cleanup() { rm -rf "${perfdata}"
@@ -154,10 +162,18 @@ test_workload() { echo "Basic target workload test [Success]" } +# raise the limit of file descriptors to minimum +if [[ $default_fd_limit -lt $min_fd_limit ]]; then + ulimit -n $min_fd_limit +fi + test_per_thread test_register_capture test_system_wide test_workload +# restore the default value +ulimit -n $default_fd_limit + cleanup exit $err
Signature · 1 line
2.41.0
Hello Arnaldo,
kind reminder, any chance this being merged soon?
Thanks,
Michael
Quoted 6 lines · On Mon, 11 Mar 2024, [email protected] wrote:
From: Veronika Molnarova <[email protected]>
Subtest for system-wide record with '--threads=cpu' option fails due to a limit of open file descriptors on systems with 128 or more CPUs as the default limit is set to 1024.
Quoted 6 lines
The number of open file descriptors should be slightly above nmb_events*nmb_cpus +
nmb_cpus(for perf.data.n)+ 4*nmb_cpus(for pipes), which equals 8*nmb_cpus. Therefore, temporarily raise the limit to 16*nmb_cpus for the test.Signed-off-by Veronika Molnarova <[email protected]>
Quoted 46 lines
--- tools/perf/tests/shell/record.sh | 16 ++++++++++++++++ 1 file changed, 16 insertions(+)tools/perf/tests/shell/record.sh +16
-0@@ -11,6 +11,14 @@ err=0 perfdata=$(mktemp /tmp/__perf_test.perf.data.XXXXX) testprog="perf test -w thloop" testsym="test_loop" +default_fd_limit=$(ulimit -n) +# With option --threads=cpu the number of open file descriptors should be +# equal to sum of: nmb_cpus * nmb_events (2+dummy), +# nmb_threads for perf.data.n (equal to nmb_cpus) and +# 2*nmb_cpus of pipes = 4*nmb_cpus (each pipe has 2 ends) +# All together it needs 8*nmb_cpus file descriptors plus some are also used +# outside of testing, thus raising the limit to 16*nmb_cpus +min_fd_limit=$(($(getconf _NPROCESSORS_ONLN) * 16)) cleanup() { rm -rf "${perfdata}"@@ -154,10 +162,18 @@ test_workload() { echo "Basic target workload test [Success]" } +# raise the limit of file descriptors to minimum +if [[ $default_fd_limit -lt $min_fd_limit ]]; then + ulimit -n $min_fd_limit +fi + test_per_thread test_register_capture test_system_wide test_workload +# restore the default value +ulimit -n $default_fd_limit + cleanup exit $errSignature · 1 line
2.41.0
Arnaldo Carvalho de Melo
maintainerQuoted 12 lines · Michael Petlan
Hello Arnaldo,
kind reminder, any chance this being merged soon?
Thanks,
MichaelQuoted 6 lines · On Mon, 11 Mar 2024, [email protected] wrote:
From: Veronika Molnarova <[email protected]>
Subtest for system-wide record with '--threads=cpu' option fails due to a limit of open file descriptors on systems with 128 or more CPUs as the default limit is set to 1024.
Quoted 8 lines
Quoted 6 lines
The number of open file descriptors should be slightly above nmb_events*nmb_cpus +
nmb_cpus(for perf.data.n)+ 4*nmb_cpus(for pipes), which equals 8*nmb_cpus. Therefore, temporarily raise the limit to 16*nmb_cpus for the test.Signed-off-by Veronika Molnarova <[email protected]>Acked-by Michael Petlan <[email protected]>
Its not applying, can you please check?
Please use thegit://git.kernel.org/pub/scm/linux/kernel/git/perf/perf-tools-next.git
repo, tmp.perf-tools-next branch.
- Arnaldo
⬢[acme@toolbox perf-tools-next]$ b4 am -ctsl --cc-trailers alpine.LRH.2.20.2403201310220.4040@Diego Grabbing thread from lore.kernel.org/all/alpine.LRH.2.20.2403201310220.4040@Diego/t.mbox.gz Checking for newer revisions
Grabbing search results from lore.kernel.org
Analyzing 2 messages in the thread
Checking attestation on all messages, may take a moment...
---
[PATCH v2] perf test record.sh: Raise limit of open file descriptors
+ Acked-by: Michael Petlan <[email protected]>
+ Link: https://lore.kernel.org/r/[email protected]
+ Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
---
NOTE: install dkimpy for DKIM signature verification
---
Total patches: 1
---
Base: not specified
git am ./v2_20240311_vmolnaro_perf_test_record_sh_raise_limit_of_open_file_descriptors.mbx
⬢[acme@toolbox perf-tools-next]$ vim ./v2_20240311_vmolnaro_perf_test_record_sh_raise_limit_of_open_file_descriptors.mbx ⬢[acme@toolbox perf-tools-next]$ patch -p1 < ./v2_20240311_vmolnaro_perf_test_record_sh_raise_limit_of_open_file_descriptors.mbx patching file tools/perf/tests/shell/record.sh
Hunk #1 succeeded at 12 with fuzz 2 (offset 1 line).
Hunk #2 FAILED at 162.
1 out of 2 hunks FAILED -- saving rejects to file tools/perf/tests/shell/record.sh.rej
⬢[acme@toolbox perf-tools-next]$
PATCH 1/1 perf test record.sh: Raise limit of open file descriptors
From: Veronika Molnarova <[email protected]>
Subtest for system-wide record with '--threads=cpu' option fails due to a limit of open file descriptors on systems with 128 or more CPUs as the default limit is set to 1024.
The number of open file descriptors should be slightly above nmb_events*nmb_cpus + nmb_cpus(for perf.data.n) + 4*nmb_cpus(for pipes), which equals 8*nmb_cpus. Therefore, temporarily raise the limit to 16*nmb_cpus for the test.
---
Reworked the patch as the testfile has already changed since the patch was sent causing the applying to fail. Should be good now.
tools/perf/tests/shell/record.sh | 16 ++++++++++++++++ 1 file changed, 16 insertions(+)
tools/perf/tests/shell/record.sh +16 -0
@@ -11,6 +11,14 @@ err=0 perfdata=$(mktemp /tmp/__perf_test.perf.data.XXXXX) testprog="perf test -w thloop" testsym="test_loop" +default_fd_limit=$(ulimit -n) +# With option --threads=cpu the number of open file descriptors should be +# equal to sum of: nmb_cpus * nmb_events (2+dummy), +# nmb_threads for perf.data.n (equal to nmb_cpus) and +# 2*nmb_cpus of pipes = 4*nmb_cpus (each pipe has 2 ends) +# All together it needs 8*nmb_cpus file descriptors plus some are also used +# outside of testing, thus raising the limit to 16*nmb_cpus +min_fd_limit=$(($(getconf _NPROCESSORS_ONLN) * 16)) cleanup() { rm -rf "${perfdata}"
@@ -154,10 +162,18 @@ test_workload() { echo "Basic target workload test [Success]" } +# raise the limit of file descriptors to minimum +if [[ $default_fd_limit -lt $min_fd_limit ]]; then + ulimit -n $min_fd_limit +fi + test_per_thread test_register_capture test_system_wide test_workload +# restore the default value +ulimit -n $default_fd_limit + cleanup exit $err
Signature · 1 line
2.43.0
Arnaldo Carvalho de Melo
maintainerQuoted 20 lines · Arnaldo Carvalho de Melo
Quoted 19 lines · Michael Petlan
Hello Arnaldo,
kind reminder, any chance this being merged soon?
Thanks,
MichaelQuoted 12 lines · On Mon, 11 Mar 2024, [email protected] wrote:
From: Veronika Molnarova <[email protected]>
Subtest for system-wide record with '--threads=cpu' option fails due to a limit of open file descriptors on systems with 128 or more CPUs as the default limit is set to 1024.
The number of open file descriptors should be slightly above nmb_events*nmb_cpus +
nmb_cpus(for perf.data.n)+ 4*nmb_cpus(for pipes), which equals 8*nmb_cpus. Therefore, temporarily raise the limit to 16*nmb_cpus for the test.Signed-off-by Veronika Molnarova <[email protected]>
Quoted 7 lines
Quoted 1 line
Acked-by Michael Petlan <[email protected]>Its not applying, can you please check?
Please use the
git://git.kernel.org/pub/scm/linux/kernel/git/perf/perf-tools-next.git
repo, tmp.perf-tools-next branch.
ping?
- Arnaldo
Quoted 28 lines
- Arnaldo
⬢[acme@toolbox perf-tools-next]$ b4 am -ctsl --cc-trailers alpine.LRH.2.20.2403201310220.4040@Diego Grabbing thread from lore.kernel.org/all/alpine.LRH.2.20.2403201310220.4040@Diego/t.mbox.gz Checking for newer revisions
Grabbing search results from lore.kernel.org
Analyzing 2 messages in the thread
Checking attestation on all messages, may take a moment...
---
[PATCH v2] perf testrecord.sh: Raise limit of open file descriptors
+ Acked-by: Michael Petlan <[email protected]>
+ Link: https://lore.kernel.org/r/[email protected]
+ Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
---
NOTE: install dkimpy for DKIM signature verification
---
Total patches: 1
---Base: not specified
git am ./v2_20240311_vmolnaro_perf_test_record_sh_raise_limit_of_open_file_descriptors.mbx⬢[acme@toolbox perf-tools-next]$ vim ./v2_20240311_vmolnaro_perf_test_record_sh_raise_limit_of_open_file_descriptors.mbx ⬢[acme@toolbox perf-tools-next]$ patch -p1 < ./v2_20240311_vmolnaro_perf_test_record_sh_raise_limit_of_open_file_descriptors.mbx patching file
tools/perf/tests/shell/record.sh
Hunk #1 succeeded at 12 with fuzz 2 (offset 1 line).
Hunk #2 FAILED at 162.
1 out of 2 hunks FAILED -- saving rejects to filetools/perf/tests/shell/record.sh.rej
⬢[acme@toolbox perf-tools-next]$
PATCH v2 1/1 perf test record.sh: Raise limit of open file descriptors
From: Veronika Molnarova <[email protected]>
Subtest for system-wide record with '--threads=cpu' option fails due to a limit of open file descriptors on systems with 128 or more CPUs as the default limit is set to 1024.
The number of open file descriptors should be slightly above nmb_events*nmb_cpus + nmb_cpus(for perf.data.n) + 4*nmb_cpus(for pipes), which equals 8*nmb_cpus. Therefore, temporarily raise the limit to 16*nmb_cpus for the test.
---
Reworked the patch as the testfile has already changed since the patch was sent causing the applying to fail, now applying upstream without problems.
tools/perf/tests/shell/record.sh | 17 +++++++++++++++++ 1 file changed, 17 insertions(+)
tools/perf/tests/shell/record.sh +17 -0
@@ -21,6 +21,15 @@ testprog="perf test -w thloop" cpu_pmu_dir="/sys/bus/event_source/devices/cpu*" br_cntr_file="/caps/branch_counter_nr" br_cntr_output="branch stack counters" +default_fd_limit=$(ulimit -n) +# With option --threads=cpu the number of open file descriptors should be +# equal to sum of: nmb_cpus * nmb_events (2+dummy), +# nmb_threads for perf.data.n (equal to nmb_cpus) and +# 2*nmb_cpus of pipes = 4*nmb_cpus (each pipe has 2 ends) +# All together it needs 8*nmb_cpus file descriptors plus some are also used +# outside of testing, thus raising the limit to 16*nmb_cpus +min_fd_limit=$(($(getconf _NPROCESSORS_ONLN) * 16)) + cleanup() { rm -rf "${perfdata}"
@@ -190,11 +199,19 @@ test_branch_counter() { echo "Basic branch counter test [Success]" } +# raise the limit of file descriptors to minimum +if [[ $default_fd_limit -lt $min_fd_limit ]]; then + ulimit -n $min_fd_limit +fi + test_per_thread test_register_capture test_system_wide test_workload test_branch_counter +# restore the default value +ulimit -n $default_fd_limit + cleanup exit $err
Signature · 1 line
2.43.0
Quoted 6 lines · On Mon, 29 Apr 2024, [email protected] wrote:
From: Veronika Molnarova <[email protected]>
Subtest for system-wide record with '--threads=cpu' option fails due to a limit of open file descriptors on systems with 128 or more CPUs as the default limit is set to 1024.
Quoted 6 lines
The number of open file descriptors should be slightly above nmb_events*nmb_cpus +
nmb_cpus(for perf.data.n)+ 4*nmb_cpus(for pipes), which equals 8*nmb_cpus. Therefore, temporarily raise the limit to 16*nmb_cpus for the test.Signed-off-by Veronika Molnarova <[email protected]>
Quoted 51 lines
---
Reworked the patch as the testfile has already changed since the patch was sent causing the applying to fail, now applying upstream without problems.tools/perf/tests/shell/record.sh | 17 +++++++++++++++++ 1 file changed, 17 insertions(+)tools/perf/tests/shell/record.sh +17
-0@@ -21,6 +21,15 @@ testprog="perf test -w thloop" cpu_pmu_dir="/sys/bus/event_source/devices/cpu*" br_cntr_file="/caps/branch_counter_nr" br_cntr_output="branch stack counters" +default_fd_limit=$(ulimit -n) +# With option --threads=cpu the number of open file descriptors should be +# equal to sum of: nmb_cpus * nmb_events (2+dummy), +# nmb_threads for perf.data.n (equal to nmb_cpus) and +# 2*nmb_cpus of pipes = 4*nmb_cpus (each pipe has 2 ends) +# All together it needs 8*nmb_cpus file descriptors plus some are also used +# outside of testing, thus raising the limit to 16*nmb_cpus +min_fd_limit=$(($(getconf _NPROCESSORS_ONLN) * 16)) + cleanup() { rm -rf "${perfdata}"@@ -190,11 +199,19 @@ test_branch_counter() { echo "Basic branch counter test [Success]" } +# raise the limit of file descriptors to minimum +if [[ $default_fd_limit -lt $min_fd_limit ]]; then + ulimit -n $min_fd_limit +fi + test_per_thread test_register_capture test_system_wide test_workload test_branch_counter +# restore the default value +ulimit -n $default_fd_limit + cleanup exit $errSignature · 1 line
2.43.0
Arnaldo Carvalho de Melo
maintainerQuoted 7 lines · Michael Petlan
Quoted 6 lines · On Mon, 29 Apr 2024, [email protected] wrote:
From: Veronika Molnarova <[email protected]>
Subtest for system-wide record with '--threads=cpu' option fails due to a limit of open file descriptors on systems with 128 or more CPUs as the default limit is set to 1024.
Quoted 8 lines
Quoted 6 lines
The number of open file descriptors should be slightly above nmb_events*nmb_cpus +
nmb_cpus(for perf.data.n)+ 4*nmb_cpus(for pipes), which equals 8*nmb_cpus. Therefore, temporarily raise the limit to 16*nmb_cpus for the test.Signed-off-by Veronika Molnarova <[email protected]>Acked-by Michael Petlan <[email protected]>
b4 got confused, I'll pick the rigth version manually...
⬢[acme@toolbox perf-tools-next]$ b4 am -ctsl --cc-trailers [email protected] Grabbing thread from lore.kernel.org/all/[email protected]/t.mbox.gz Checking for newer revisions
Grabbing search results from lore.kernel.org
Analyzing 7 messages in the thread
WARNING: duplicate messages found at index 1
Subject 1: perf test record.sh: Raise limit of open file descriptors
Subject 2: perf test record.sh: Raise limit of open file descriptors
2 is not a reply... assume additional patch
Looking for additional code-review trailers on lore.kernel.org
Will use the latest revision: v2
You can pick other revisions using the -vN flag
Checking attestation on all messages, may take a moment...
---
✓ [PATCH v2] perf test record.sh: Raise limit of open file descriptors
+ Link: https://lore.kernel.org/r/[email protected] (✓ DKIM/kernel.org)
+ Acked-by: Michael Petlan <[email protected]> (✓ DKIM/redhat.com)
+ Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
✓ [PATCH v2] perf test record.sh: Raise limit of open file descriptors
+ Acked-by: Michael Petlan <[email protected]> (✓ DKIM/redhat.com)
+ Link: https://lore.kernel.org/r/[email protected]
+ Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
---
✓ Signed: DKIM/redhat.com
---
Total patches: 2
---
Base: not specified
git am ./v2_20240311_vmolnaro_perf_test_record_sh_raise_limit_of_open_file_descriptors.mbx
⬢[acme@toolbox perf-tools-next]$
Quoted 52 lines
Quoted 51 lines
---
Reworked the patch as the testfile has already changed since the patch was sent causing the applying to fail, now applying upstream without problems.tools/perf/tests/shell/record.sh | 17 +++++++++++++++++ 1 file changed, 17 insertions(+)tools/perf/tests/shell/record.sh +17
-0@@ -21,6 +21,15 @@ testprog="perf test -w thloop" cpu_pmu_dir="/sys/bus/event_source/devices/cpu*" br_cntr_file="/caps/branch_counter_nr" br_cntr_output="branch stack counters" +default_fd_limit=$(ulimit -n) +# With option --threads=cpu the number of open file descriptors should be +# equal to sum of: nmb_cpus * nmb_events (2+dummy), +# nmb_threads for perf.data.n (equal to nmb_cpus) and +# 2*nmb_cpus of pipes = 4*nmb_cpus (each pipe has 2 ends) +# All together it needs 8*nmb_cpus file descriptors plus some are also used +# outside of testing, thus raising the limit to 16*nmb_cpus +min_fd_limit=$(($(getconf _NPROCESSORS_ONLN) * 16)) + cleanup() { rm -rf "${perfdata}"@@ -190,11 +199,19 @@ test_branch_counter() { echo "Basic branch counter test [Success]" } +# raise the limit of file descriptors to minimum +if [[ $default_fd_limit -lt $min_fd_limit ]]; then + ulimit -n $min_fd_limit +fi + test_per_thread test_register_capture test_system_wide test_workload test_branch_counter +# restore the default value +ulimit -n $default_fd_limit + cleanup exit $errSignature · 1 line
2.43.0
Arnaldo Carvalho de Melo
maintainerQuoted 13 lines · Arnaldo Carvalho de Melo
Quoted 12 lines · Michael Petlan
Quoted 11 lines · On Mon, 29 Apr 2024, [email protected] wrote:
From: Veronika Molnarova <[email protected]>
Subtest for system-wide record with '--threads=cpu' option fails due to a limit of open file descriptors on systems with 128 or more CPUs as the default limit is set to 1024.
The number of open file descriptors should be slightly above nmb_events*nmb_cpus +
nmb_cpus(for perf.data.n)+ 4*nmb_cpus(for pipes), which equals 8*nmb_cpus. Therefore, temporarily raise the limit to 16*nmb_cpus for the test.
Quoted 5 lines
Quoted 3 lines
Quoted 1 line
Signed-off-by Veronika Molnarova <[email protected]>Acked-by Michael Petlan <[email protected]>b4 got confused, I'll pick the rigth version manually...
And you forgot to add the perf tools reviewers, as listed in MAINTAINERS, you can get the list of people who should be on the Cc using:
⬢[acme@toolbox perf-tools-next]$ scripts/get_maintainer.pl tools/perf/tests/shell/record.sh
Peter Zijlstra <[email protected]> (supporter:PERFORMANCE EVENTS SUBSYSTEM)
Ingo Molnar <[email protected]> (supporter:PERFORMANCE EVENTS SUBSYSTEM)
Arnaldo Carvalho de Melo <[email protected]> (supporter:PERFORMANCE EVENTS SUBSYSTEM,commit_signer:3/4=75%)
Namhyung Kim <[email protected]> (supporter:PERFORMANCE EVENTS SUBSYSTEM,commit_signer:1/4=25%)
Mark Rutland <[email protected]> (reviewer:PERFORMANCE EVENTS SUBSYSTEM)
Alexander Shishkin <[email protected]> (reviewer:PERFORMANCE EVENTS SUBSYSTEM)
Jiri Olsa <[email protected]> (reviewer:PERFORMANCE EVENTS SUBSYSTEM)
Ian Rogers <[email protected]> (reviewer:PERFORMANCE EVENTS SUBSYSTEM)
Adrian Hunter <[email protected]> (reviewer:PERFORMANCE EVENTS SUBSYSTEM,commit_signer:1/4=25%,authored:1/4=25%,added_lines:7/55=13%,removed_lines:1/1=100%) "Liang, Kan" <[email protected]> (reviewer:PERFORMANCE EVENTS SUBSYSTEM,commit_signer:1/4=25%,authored:1/4=25%,added_lines:30/55=55%) Athira Rajeev <[email protected]> (commit_signer:1/4=25%,authored:1/4=25%)
Veronika Molnarova <[email protected]> (authored:1/4=25%,added_lines:17/55=31%)
[email protected] (open list:PERFORMANCE EVENTS SUBSYSTEM)
[email protected] (open list:PERFORMANCE EVENTS SUBSYSTEM)
⬢[acme@toolbox perf-tools-next]$
Quoted 84 lines
⬢[acme@toolbox perf-tools-next]$ b4 am -ctsl --cc-trailers [email protected] Grabbing thread from lore.kernel.org/all/[email protected]/t.mbox.gz Checking for newer revisions
Grabbing search results from lore.kernel.org
Analyzing 7 messages in the thread
WARNING: duplicate messages found at index 1
Subject 1: perf testrecord.sh: Raise limit of open file descriptors
Subject 2: perf testrecord.sh: Raise limit of open file descriptors
2 is not a reply... assume additional patch
Looking for additional code-review trailers on lore.kernel.org
Will use the latest revision: v2
You can pick other revisions using the -vN flag
Checking attestation on all messages, may take a moment...
---
✓ [PATCH v2] perf testrecord.sh: Raise limit of open file descriptors
+ Link: https://lore.kernel.org/r/[email protected] (✓ DKIM/kernel.org)
+ Acked-by: Michael Petlan <[email protected]> (✓ DKIM/redhat.com)
+ Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
✓ [PATCH v2] perf testrecord.sh: Raise limit of open file descriptors
+ Acked-by: Michael Petlan <[email protected]> (✓ DKIM/redhat.com)
+ Link: https://lore.kernel.org/r/[email protected]
+ Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
---
✓ Signed: DKIM/redhat.com
---
Total patches: 2
---Base: not specified
git am ./v2_20240311_vmolnaro_perf_test_record_sh_raise_limit_of_open_file_descriptors.mbx⬢[acme@toolbox perf-tools-next]$
Quoted 52 lines
Quoted 51 lines
---
Reworked the patch as the testfile has already changed since the patch was sent causing the applying to fail, now applying upstream without problems.tools/perf/tests/shell/record.sh | 17 +++++++++++++++++ 1 file changed, 17 insertions(+)tools/perf/tests/shell/record.sh +17
-0@@ -21,6 +21,15 @@ testprog="perf test -w thloop" cpu_pmu_dir="/sys/bus/event_source/devices/cpu*" br_cntr_file="/caps/branch_counter_nr" br_cntr_output="branch stack counters" +default_fd_limit=$(ulimit -n) +# With option --threads=cpu the number of open file descriptors should be +# equal to sum of: nmb_cpus * nmb_events (2+dummy), +# nmb_threads for perf.data.n (equal to nmb_cpus) and +# 2*nmb_cpus of pipes = 4*nmb_cpus (each pipe has 2 ends) +# All together it needs 8*nmb_cpus file descriptors plus some are also used +# outside of testing, thus raising the limit to 16*nmb_cpus +min_fd_limit=$(($(getconf _NPROCESSORS_ONLN) * 16)) + cleanup() { rm -rf "${perfdata}"@@ -190,11 +199,19 @@ test_branch_counter() { echo "Basic branch counter test [Success]" } +# raise the limit of file descriptors to minimum +if [[ $default_fd_limit -lt $min_fd_limit ]]; then + ulimit -n $min_fd_limit +fi + test_per_thread test_register_capture test_system_wide test_workload test_branch_counter +# restore the default value +ulimit -n $default_fd_limit + cleanup exit $errSignature · 1 line
2.43.0
Arnaldo Carvalho de Melo
maintainerQuoted 14 lines · Arnaldo Carvalho de Melo
Quoted 12 lines · Arnaldo Carvalho de Melo
Quoted 11 lines · Michael Petlan
Quoted 10 lines · On Mon, 29 Apr 2024, [email protected] wrote:
From: Veronika Molnarova <[email protected]>
Subtest for system-wide record with '--threads=cpu' option fails due to a limit of open file descriptors on systems with 128 or more CPUs as the default limit is set to 1024.
The number of open file descriptors should be slightly above nmb_events*nmb_cpus +
nmb_cpus(for perf.data.n)+ 4*nmb_cpus(for pipes), which equals 8*nmb_cpus. Therefore, temporarily raise the limit to 16*nmb_cpus for the test.
Quoted 5 lines
Quoted 5 lines
Quoted 3 lines
Quoted 1 line
Signed-off-by Veronika Molnarova <[email protected]>Acked-by Michael Petlan <[email protected]>b4 got confused, I'll pick the rigth version manually...
Also please install ShellCheck, with it installed, as I do, you'll get this warning, that we should address:
CC /tmp/build/perf-tools-next/util/synthetic-events.o
In tests/shell/record.sh line 24:
default_fd_limit=$(ulimit -n)
^-- SC3045 (warning): In POSIX sh, ulimit -n is undefined.
In tests/shell/record.sh line 203:
if [[ $default_fd_limit -lt $min_fd_limit ]]; then
^-- SC3010 (warning): In POSIX sh, [[ ]] is undefined.
In tests/shell/record.sh line 204:
ulimit -n $min_fd_limit
^-- SC3045 (warning): In POSIX sh, ulimit -n is undefined.
In tests/shell/record.sh line 214:
ulimit -n $default_fd_limit
^-- SC3045 (warning): In POSIX sh, ulimit -n is undefined.
For more information:
https://www.shellcheck.net/wiki/SC3010 -- In POSIX sh, [[ ]] is undefined.
https://www.shellcheck.net/wiki/SC3045 -- In POSIX sh, ulimit -n is undefin...
make[4]: *** [tests/Build:90: /tmp/build/perf-tools-next/tests/shell/record.sh.shellcheck_log] Error 1 make[3]: *** [/home/acme/git/perf-tools-next/tools/build/Makefile.build:158: tests] Error 2 make[2]: *** [Makefile.perf:777: /tmp/build/perf-tools-next/perf-test-in.o] Error 2
CC /tmp/build/perf-tools-next/util/data.o
make[2]: *** Waiting for unfinished jobs....