Re: [PATCH 2/2] perf test shell: make perf inet_pton test more portable

From: Thomas-Mich Richter
Date: Wed Jun 20 2018 - 05:15:45 EST

Next message: Peter Zijlstra: "Re: [PATCH v10 7/7] x86/tsc: use tsc early"
Previous message: Gustavo Pimentel: "Re: [PATCH v4 03/10] PCI: Update xxx_pcie_ep_raise_irq() and pci_epc_raise_irq() signatures"
In reply to: Kim Phillips: "[PATCH 2/2] perf test shell: make perf inet_pton test more portable"
Next in thread: Arnaldo Carvalho de Melo: "Re: [PATCH 2/2] perf test shell: make perf inet_pton test more portable"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

On 06/20/2018 01:49 AM, Kim Phillips wrote:
> Debian based systems such as Ubuntu have dash as their default shell.
> Even if the normal or root user's shell is bash, certain scripts still
> call /bin/sh, which points to dash, so we fix this perf test by
> rewriting it in a more portable way.
>
> BEFORE:
>
> $ sudo ./perf test -v 64
> 64: probe libc's inet_pton & backtrace it with ping :
> --- start ---
> test child forked, pid 31942
> ./tests/shell/record+probe_libc_inet_pton.sh: 18: ./tests/shell/record+probe_libc_inet_pton.sh: expected[0]=ping[][0-9 \.:]+probe_libc:inet_pton: $[[:xdigit:]]+$: not found
> ./tests/shell/record+probe_libc_inet_pton.sh: 19: ./tests/shell/record+probe_libc_inet_pton.sh: expected[1]=.*inet_pton\+0x[[:xdigit:]]+[[:space:]]$/lib/x86_64-linux-gnu/libc-2.27.so|inlined$$: not found
> ./tests/shell/record+probe_libc_inet_pton.sh: 29: ./tests/shell/record+probe_libc_inet_pton.sh: expected[2]=getaddrinfo\+0x[[:xdigit:]]+[[:space:]]$/lib/x86_64-linux-gnu/libc-2.27.so$$: not found
> ./tests/shell/record+probe_libc_inet_pton.sh: 30: ./tests/shell/record+probe_libc_inet_pton.sh: expected[3]=.*\+0x[[:xdigit:]]+[[:space:]]$.*/bin/ping.*$$: not found
> ping 31963 [004] 83577.670613: probe_libc:inet_pton: (7fe15f87f4b0)
> ./tests/shell/record+probe_libc_inet_pton.sh: 39: ./tests/shell/record+probe_libc_inet_pton.sh: Bad substitution
> ./tests/shell/record+probe_libc_inet_pton.sh: 41: ./tests/shell/record+probe_libc_inet_pton.sh: Bad substitution
> test child finished with -2
> ---- end ----
> probe libc's inet_pton & backtrace it with ping: Skip
>
> AFTER:
>
> 64: probe libc's inet_pton & backtrace it with ping :
> --- start ---
> test child forked, pid 32277
> ping 32295 [001] 83679.690020: probe_libc:inet_pton: (7ff244f504b0)
> 7ff244f504b0 __GI___inet_pton+0x0 (/lib/x86_64-linux-gnu/libc-2.27.so)
> 7ff244f14ce4 getaddrinfo+0x124 (/lib/x86_64-linux-gnu/libc-2.27.so)
> 556ac036b57d _init+0xb75 (/bin/ping)
> test child finished with 0
> ---- end ----
> probe libc's inet_pton & backtrace it with ping: Ok
>
> Cc: Arnaldo Carvalho de Melo <acme@xxxxxxxxxx>
> Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
> Cc: Ingo Molnar <mingo@xxxxxxxxxx>
> Cc: Alexander Shishkin <alexander.shishkin@xxxxxxxxxxxxxxx>
> Cc: Jiri Olsa <jolsa@xxxxxxxxxx>
> Cc: Namhyung Kim <namhyung@xxxxxxxxxx>
> Cc: Thomas Richter <tmricht@xxxxxxxxxxxxxxxxxx>
> Cc: Michael Petlan <mpetlan@xxxxxxxxxx>
> Cc: "Hendrik BrÃckner" <brueckner@xxxxxxxxxxxxxxxxxx>
> Cc: Sandipan Das <sandipan@xxxxxxxxxxxxxxxxxx>
> Signed-off-by: Kim Phillips <kim.phillips@xxxxxxx>
> ---
> Fixed using fake array technique described here:
>
> https://www.linuxquestions.org/questions/programming-9/basic-bash-how-to-use-eval-to-evaluate-variable-names-made-of-arbitrary-strings-775622/
>
> Note that bashisms can also be found with 'checkbashisms':
>
> $ checkbashisms record+probe_libc_inet_pton.sh
> script record+probe_libc_inet_pton.sh does not appear to have a #! interpreter line;
> you may get strange results
> possible bashism in record+probe_libc_inet_pton.sh line 18 (bash arrays, H[0]):
> expected[0]="ping[][0-9 \.:]+probe_libc:inet_pton: $[[:xdigit:]]+$"
> possible bashism in record+probe_libc_inet_pton.sh line 19 (bash arrays, H[0]):
> expected[1]=".*inet_pton\+0x[[:xdigit:]]+[[:space:]]$$libc|inlined$$"
> possible bashism in record+probe_libc_inet_pton.sh line 23 (bash arrays, H[0]):
> expected[2]="gaih_inet.*\+0x[[:xdigit:]]+[[:space:]]$$libc|inlined$$"
> possible bashism in record+probe_libc_inet_pton.sh line 24 (bash arrays, H[0]):
> expected[3]="(__GI_)?getaddrinfo\+0x[[:xdigit:]]+[[:space:]]$$libc|inlined$$"
> possible bashism in record+probe_libc_inet_pton.sh line 25 (bash arrays, H[0]):
> expected[4]="main\+0x[[:xdigit:]]+[[:space:]]$.*/bin/ping.*$$"
> possible bashism in record+probe_libc_inet_pton.sh line 29 (bash arrays, H[0]):
> expected[2]="getaddrinfo\+0x[[:xdigit:]]+[[:space:]]$$libc$$"
> possible bashism in record+probe_libc_inet_pton.sh line 30 (bash arrays, H[0]):
> expected[3]=".*\+0x[[:xdigit:]]+[[:space:]]$.*/bin/ping.*$$"
> possible bashism in record+probe_libc_inet_pton.sh line 39 (bash arrays, ${name[0|*|@]}):
> echo "$line" | egrep -q "${expected[$idx]}"
> possible bashism in record+probe_libc_inet_pton.sh line 41 (bash arrays, ${name[0|*|@]}):
> printf "FAIL: expected backtrace entry %d \"%s\" got \"%s\"\n" $idx "${expected[$idx]}" "$line"
> possible bashism in record+probe_libc_inet_pton.sh line 44 (let ...):
> let idx+=1
> possible bashism in record+probe_libc_inet_pton.sh line 45 (bash arrays, ${name[0|*|@]}):
> [ -z "${expected[$idx]}" ] && break
>
> .../shell/record+probe_libc_inet_pton.sh | 23 ++++++++++---------
> 1 file changed, 12 insertions(+), 11 deletions(-)
>
> diff --git a/tools/perf/tests/shell/record+probe_libc_inet_pton.sh b/tools/perf/tests/shell/record+probe_libc_inet_pton.sh
> index 263057039693..41124fa12913 100755
> --- a/tools/perf/tests/shell/record+probe_libc_inet_pton.sh
> +++ b/tools/perf/tests/shell/record+probe_libc_inet_pton.sh
> @@ -15,19 +15,20 @@ nm -Dg $libc 2>/dev/null | fgrep -q inet_pton || exit 254
>
> trace_libc_inet_pton_backtrace() {
> idx=0
> - expected[0]="ping[][0-9 \.:]+probe_libc:inet_pton: $[[:xdigit:]]+$"
> - expected[1]=".*inet_pton\+0x[[:xdigit:]]+[[:space:]]$$libc|inlined$$"
> + expected='expected'
> + eval ${expected}0="'ping[][0-9 \.:]+probe_libc:inet_pton: $[[:xdigit:]]+$'"
> + eval ${expected}1="'.*inet_pton\+0x[[:xdigit:]]+[[:space:]]$$libc|inlined$$'"
> case "$(uname -m)" in
> s390x)
> eventattr='call-graph=dwarf,max-stack=4'
> - expected[2]="gaih_inet.*\+0x[[:xdigit:]]+[[:space:]]$$libc|inlined$$"
> - expected[3]="(__GI_)?getaddrinfo\+0x[[:xdigit:]]+[[:space:]]$$libc|inlined$$"
> - expected[4]="main\+0x[[:xdigit:]]+[[:space:]]$.*/bin/ping.*$$"
> + eval ${expected}2="'gaih_inet.*\+0x[[:xdigit:]]+[[:space:]]$$libc|inlined$$'"
> + eval ${expected}3="'(__GI_)?getaddrinfo\+0x[[:xdigit:]]+[[:space:]]$$libc|inlined$$'"
> + eval ${expected}4="'main\+0x[[:xdigit:]]+[[:space:]]$.*/bin/ping.*$$'"
> ;;
> *)
> eventattr='max-stack=3'
> - expected[2]="getaddrinfo\+0x[[:xdigit:]]+[[:space:]]$$libc$$"
> - expected[3]=".*\+0x[[:xdigit:]]+[[:space:]]$.*/bin/ping.*$$"
> + eval ${expected}2="'getaddrinfo\+0x[[:xdigit:]]+[[:space:]]$$libc$$'"
> + eval ${expected}3="'.*\+0x[[:xdigit:]]+[[:space:]]$.*/bin/ping.*$$'"
> ;;
> esac
>
> @@ -36,13 +37,13 @@ trace_libc_inet_pton_backtrace() {
> perf record -e probe_libc:inet_pton/$eventattr/ -o $file ping -6 -c 1 ::1 > /dev/null 2>&1
> perf script -i $file | while read line ; do
> echo $line
> - echo "$line" | egrep -q "${expected[$idx]}"
> + echo "$line" | egrep -q "$(echo $(eval echo \$$expected${idx}))"
> if [ $? -ne 0 ] ; then
> - printf "FAIL: expected backtrace entry %d \"%s\" got \"%s\"\n" $idx "${expected[$idx]}" "$line"
> + printf "FAIL: expected backtrace entry %d \"%s\" got \"%s\"\n" $idx "$(echo $(eval echo \$$expected${idx}))" "$line"
> exit 1
> fi
> - let idx+=1
> - [ -z "${expected[$idx]}" ] && break
> + idx=$((idx+1))
> + [ -z "$(echo $(eval echo \$$expected${idx}))" ] && break
> done
>
> # If any statements are executed from this point onwards,
>

I tried it on my s390 and it works nicely.

Tested-by: Thomas Richter <tmricht@xxxxxxxxxxxxx>

--
Thomas Richter, Dept 3303, IBM s390 Linux Development, Boeblingen, Germany
--
Vorsitzende des Aufsichtsrats: Martina Koederitz
GeschÃftsfÃhrung: Dirk Wittkopp
Sitz der Gesellschaft: BÃblingen / Registergericht: Amtsgericht Stuttgart, HRB 243294

Next message: Peter Zijlstra: "Re: [PATCH v10 7/7] x86/tsc: use tsc early"
Previous message: Gustavo Pimentel: "Re: [PATCH v4 03/10] PCI: Update xxx_pcie_ep_raise_irq() and pci_epc_raise_irq() signatures"
In reply to: Kim Phillips: "[PATCH 2/2] perf test shell: make perf inet_pton test more portable"
Next in thread: Arnaldo Carvalho de Melo: "Re: [PATCH 2/2] perf test shell: make perf inet_pton test more portable"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]