shell: ROD_SILENT: Print std{out, err} on failure

RFC PATCH New Cyril Hrubis 3 participants 8 messages Feb 12, 2018 – Mar 14, 2018

Share & tools

Cyril Hrubis

author

2018-02-12 15:03 UTC #0 lore raw

RFC PATCH 1/1 shell: ROD_SILENT: Print std{out, err} on failure

Recently we had a problem when openQA (test automation framework) passed wrong device path as LTP_BIG_DEV. It has been much more complicated to debug because the error message from mkfs was not shown at all, since the tst_mkfs uses ROD_SILENT to execute the mkfs command.

The best solution for ROD_SILENT is to store the command output into a variable and print it only on failure which is done by this patch.

Signed-off-by Cyril Hrubis <[email protected]>

---
 testcases/lib/test.sh     | 3 ++-
 testcases/lib/tst_test.sh | 3 ++-
 2 files changed, 4 insertions(+), 2 deletions(-)

testcases/lib/test.sh +2 -1

index 341c1f0f4..bce9893a9 100644
@@ -225,8 +225,9 @@ tst_timeout()
 
 ROD_SILENT()
 {
-	$@ > /dev/null 2>&1
+	local tst_out=$($@ 2>&1)
 	if [ $? -ne 0 ]; then
+		echo "$tst_out"
 		tst_brkm TBROK "$@ failed"
 	fi
 }

testcases/lib/tst_test.sh +2 -1

index d6b638549..48afb9cc4 100644
@@ -118,8 +118,9 @@ tst_brk()
 
 ROD_SILENT()
 {
-	tst_rod $@ > /dev/null 2>&1
+	local tst_out=$(tst_rod $@ 2>&1)
 	if [ $? -ne 0 ]; then
+		echo "$tst_out"
 		tst_brk TBROK "$@ failed"
 	fi
 }

Signature · 1 line

2.13.6

Richard Palethorpe

2018-02-21 09:12 UTC #1 lore raw

RFC PATCH 1/1 shell: ROD_SILENT: Print std{out, err} on failure

Hello,

Quoted 46 lines · Cyril Hrubis

Recently we had a problem when openQA (test automation framework) passed wrong device path as LTP_BIG_DEV. It has been much more complicated to debug because the error message from mkfs was not shown at all, since the tst_mkfs uses ROD_SILENT to execute the mkfs command.
The best solution for ROD_SILENT is to store the command output into a variable and print it only on failure which is done by this patch.
Signed-off-by Cyril Hrubis <[email protected]>
---
 testcases/lib/test.sh     | 3 ++-
 testcases/lib/tst_test.sh | 3 ++-
 2 files changed, 4 insertions(+), 2 deletions(-)
testcases/lib/test.sh +2 -1
index 341c1f0f4..bce9893a9 100644
@@ -225,8 +225,9 @@ tst_timeout()
 
 ROD_SILENT()
 {
-	$@ > /dev/null 2>&1
+	local tst_out=$($@ 2>&1)
 	if [ $? -ne 0 ]; then
+		echo "$tst_out"
 		tst_brkm TBROK "$@ failed"
 	fi
 }
testcases/lib/tst_test.sh +2 -1
index d6b638549..48afb9cc4 100644
@@ -118,8 +118,9 @@ tst_brk()
 
 ROD_SILENT()
 {
-	tst_rod $@ > /dev/null 2>&1
+	local tst_out=$(tst_rod $@ 2>&1)
 	if [ $? -ne 0 ]; then
+		echo "$tst_out"
 		tst_brk TBROK "$@ failed"
 	fi
 }
Signature · 1 line
2.13.6

LGTM except that you could maybe use ${ } instead of $( ) as a small optimisation.

Signature · 2 lines

Thank you,
Richard.

Li Wang

2018-02-22 04:05 UTC #2 lore raw

RFC PATCH 1/1 shell: ROD_SILENT: Print std{out, err} on failure

Quoted 39 lines · Richard Palethorpe

Hello,
Quoted 35 lines · Cyril Hrubis
Recently we had a problem when openQA (test automation framework) passed wrong device path as LTP_BIG_DEV. It has been much more complicated to debug because the error message from mkfs was not shown at all, since the tst_mkfs uses ROD_SILENT to execute the mkfs command.
The best solution for ROD_SILENT is to store the command output into a variable and print it only on failure which is done by this patch.
Signed-off-by Cyril Hrubis <[email protected]>
---
 testcases/lib/test.sh     | 3 ++-
 testcases/lib/tst_test.sh | 3 ++-
 2 files changed, 4 insertions(+), 2 deletions(-)
testcases/lib/test.sh +2 -1
index 341c1f0f4..bce9893a9 100644
@@ -225,8 +225,9 @@ tst_timeout()
 
 ROD_SILENT()
 {
-     $@ > /dev/null 2>&1
+     local tst_out=$($@ 2>&1)
      if [ $? -ne 0 ]; then
+             echo "$tst_out"
              tst_brkm TBROK "$@ failed"
      fi
 }
testcases/lib/tst_test.sh +0 -0
index d6b638549..48afb9cc4 100644
@@ -118,8 +118,9 @@ tst_brk()
 

Quoted 15 lines

Quoted 11 lines
testcases/lib/tst_test.sh +2 -2
 ROD_SILENT()
 {
-     tst_rod $@ > /dev/null 2>&1
+     local tst_out=$(tst_rod $@ 2>&1)
      if [ $? -ne 0 ]; then
+             echo "$tst_out"
              tst_brk TBROK "$@ failed"
      fi
 }
--
2.13.6
LGTM except that you could maybe use ${ } instead of $( ) as a small optimisation.

For
what optimisation?

$(command) runs the command and captures its output in separate shell. ${parameter} the value of parameter is substituted.

Here the purpose is to hide the output in sub shell I guess, shouldn't use $()?

Signature · 5 lines

Li Wang
[email protected]
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linux.it/pipermail/ltp/attachments/20180222/36cdf716/attachment.html>

Cyril Hrubis

author

2018-03-13 12:46 UTC #3 lore raw

RFC PATCH 1/1 shell: ROD_SILENT: Print std{out, err} on failure

In reply to Richard Palethorpe

Hi!

Quoted 6 lines

Quoted 6 lines
testcases/lib/tst_test.sh +0 -0
index d6b638549..48afb9cc4 100644
@@ -118,8 +118,9 @@ tst_brk()
 

Quoted 14 lines

Quoted 11 lines
testcases/lib/tst_test.sh +2 -1
 ROD_SILENT()
 {
-	tst_rod $@ > /dev/null 2>&1
+	local tst_out=$(tst_rod $@ 2>&1)
 	if [ $? -ne 0 ]; then
+		echo "$tst_out"
 		tst_brk TBROK "$@ failed"
 	fi
 }
Signature · 1 line
2.13.6
LGTM except that you could maybe use ${ } instead of $( ) as a small optimisation.

I do not get it either. Where should I put the ${} insetad of $()?

Signature · 2 lines

Cyril Hrubis
[email protected]

Richard Palethorpe

2018-03-13 13:01 UTC #4 lore raw

RFC PATCH 1/1 shell: ROD_SILENT: Print std{out, err} on failure

Hello,

Quoted 22 lines · Cyril Hrubis

Hi!
Quoted 6 lines
Quoted 6 lines
testcases/lib/tst_test.sh +0 -0
index d6b638549..48afb9cc4 100644
@@ -118,8 +118,9 @@ tst_brk()
 
Quoted 14 lines
Quoted 11 lines
testcases/lib/tst_test.sh +2 -2
 ROD_SILENT()
 {
-	tst_rod $@ > /dev/null 2>&1
+	local tst_out=$(tst_rod $@ 2>&1)
 	if [ $? -ne 0 ]; then
+		echo "$tst_out"
 		tst_brk TBROK "$@ failed"
 	fi
 }
--
2.13.6
LGTM except that you could maybe use ${ } instead of $( ) as a small optimisation.

Quoted 1 line

I do not get it either. Where should I put the ${} insetad of $()?

Sorry I was probably wrong, unless you can somehow change from variable substitution to redirecting from a group (that is {} > $var not var=${}).

--
Thank you,
Richard.

Cyril Hrubis

author

2018-03-13 13:30 UTC #5 lore raw

RFC PATCH 1/1 shell: ROD_SILENT: Print std{out, err} on failure

Hi!

Quoted 8 lines

Quoted 4 lines
Quoted 2 lines
LGTM except that you could maybe use ${ } instead of $( ) as a small optimisation.
I do not get it either. Where should I put the ${} insetad of $()?
Sorry I was probably wrong, unless you can somehow change from variable substitution to redirecting from a group (that is {} > $var not var=${}).

So can I consider the original patch acked?

Signature · 2 lines

Cyril Hrubis
[email protected]

Richard Palethorpe

2018-03-13 15:16 UTC #6 lore raw

RFC PATCH 1/1 shell: ROD_SILENT: Print std{out, err} on failure

Quoted 11 lines · Cyril Hrubis

Hi!
Quoted 8 lines
Quoted 4 lines
Quoted 2 lines
LGTM except that you could maybe use ${ } instead of $( ) as a small optimisation.
I do not get it either. Where should I put the ${} insetad of $()?
Sorry I was probably wrong, unless you can somehow change from variable substitution to redirecting from a group (that is {} > $var not var=${}).
So can I consider the original patch acked?

Yes.

Signature · 2 lines

Thank you,
Richard.

Cyril Hrubis

author

2018-03-14 13:58 UTC #7 lore raw

RFC PATCH 1/1 shell: ROD_SILENT: Print std{out, err} on failure

Hi!

Quoted 1 line

Yes.

Pushed, thanks.

Signature · 2 lines

Cyril Hrubis
[email protected]