[PATCH] test-lib: limit the output of the yes utility

Johannes Sixt· Feb 2, 2016, 07:28 UTC · lore

On Windows, there is no SIGPIPE. A consequence of this is that the upstream process of a pipe does not notice the death of the downstream process until the pipe buffer is full and writing more data returns an error. This behavior is the reason for an annoying delay during the execution of t7610-mergetool.sh: There are a number of test cases where 'yes' is invoked upstream. Since the utility is basically an endless loop it runs, on Windows, until the pipe buffer is full. This does take a few seconds.

The test suite has its own implementation of 'yes'. Modify it to produce only a limited amount of output that is sufficient for the test suite. The amount chosen should be sufficiently high for any test case, assuming that future test cases will not exaggerate their demands of input from an upstream 'yes' invocation.

Signed-off-by: Johannes Sixt <j6t@kdbg.org>
---
 This does not fix an error, but only an unnecessary sink of CPU cycles
 and wasted wall clock time.

 t/test-lib.sh | 14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)

Show changes to t/test-lib.sh +7 −7

diff --git a/t/test-lib.sh b/t/test-lib.sh
index bd4b02e..97e6491 100644
--- a/t/test-lib.sh
+++ b/t/test-lib.sh
@@ -902,15 +902,15 @@ fi
 yes () {
 	if test $# = 0
 	then
-		y=y
+		set -- y
 	else
-		y="$*"
+		set -- "$*"
 	fi
-
-	while echo "$y"
-	do
-		:
-	done
+	# we do not need an infinite supply of output for tests
+	set -- "$@" "$@" "$@" "$@"	# 4
+	set -- "$@" "$@" "$@" "$@"	# 16
+	set -- "$@" "$@" "$@" "$@"	# 64
+	printf "%s\n" "$@"
 }
 
 # Fix some commands on Windows

-- 
2.7.0.118.g90056ae

Johannes Schindelin· Feb 2, 2016, 08:21 UTC · re: Johannes Sixt · lore

Re: [PATCH] test-lib: limit the output of the yes utility

Hi Hannes,

On Tue, 2 Feb 2016, Johannes Sixt wrote:

> On Windows, there is no SIGPIPE.

True. But we do get some sort of write failure, no? Otherwise https://github.com/git/git/commit/2b86292ed would not work...

Show 25 quoted lines

> diff --git a/t/test-lib.sh b/t/test-lib.sh
> index bd4b02e..97e6491 100644
> --- a/t/test-lib.sh
> +++ b/t/test-lib.sh
> @@ -902,15 +902,15 @@ fi
>  yes () {
>  	if test $# = 0
>  	then
> -		y=y
> +		set -- y
>  	else
> -		y="$*"
> +		set -- "$*"
>  	fi
> -
> -	while echo "$y"
> -	do
> -		:
> -	done
> +	# we do not need an infinite supply of output for tests
> +	set -- "$@" "$@" "$@" "$@"	# 4
> +	set -- "$@" "$@" "$@" "$@"	# 16
> +	set -- "$@" "$@" "$@" "$@"	# 64
> +	printf "%s\n" "$@"
>  }

I agree with the idea, but I would like to have a less intrusive patch. Something like this should do the job as well, and is a little easier on the eyes IMHO:

-- snipsnap --

Show changes to t/test-lib.sh +4 −2

diff --git a/t/test-lib.sh b/t/test-lib.sh
index 32ac1a6..ae381b9 100644
--- a/t/test-lib.sh
+++ b/t/test-lib.sh
@@ -907,9 +907,11 @@ yes () {
 		y="$*"
 	fi
 
-	while echo "$y"
+	i=0
+	while test $i -lt 99
 	do
-		:
+		echo "$y"
+		i=$(($i+1))
 	done
 }

Johannes Sixt· Feb 2, 2016, 18:15 UTC · re: Johannes Schindelin · lore

[PATCH v2] test-lib: limit the output of the yes utility

From: Johannes Schindelin <johannes.schindelin@gmx.de>

On Windows, there is no SIGPIPE. A consequence of this is that the upstream process of a pipe does not notice the death of the downstream process until the pipe buffer is full and writing more data returns an error. This behavior is the reason for an annoying delay during the execution of t7610-mergetool.sh: There are a number of test cases where 'yes' is invoked upstream. Since the utility is basically an endless loop it runs, on Windows, until the pipe buffer is full. This does take a few seconds.

The test suite has its own implementation of 'yes'. Modify it to produce only a limited amount of output that is sufficient for the test suite. The amount chosen should be sufficiently high for any test case, assuming that future test cases will not exaggerate their demands of input from an upstream 'yes' invocation.

[j6t: commit message]

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Johannes Sixt <j6t@kdbg.org>
---
Am 02.02.2016 um 09:21 schrieb Johannes Schindelin:

Show 5 quoted lines

> On Tue, 2 Feb 2016, Johannes Sixt wrote:
>> On Windows, there is no SIGPIPE.
> 
> True. But we do get some sort of write failure, no? Otherwise
> https://github.com/git/git/commit/2b86292ed would not work...

Of course. See the second sentence of the commit message.

> I agree with the idea, but I would like to have a less intrusive patch.
> Something like this should do the job as well, and is a little easier on
> the eyes IMHO:

I'm not 100% satisfied with your version because it stomps on short-and-sweet shell variables $i and $y. But then the utility only occurs upstream of a pipeline in a separate process, so that does not matter.

Please allow me to pass authorship to you, since the patch text is now all yours, and to forge your sign-off.

-- Hannes

 t/test-lib.sh | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

Show changes to t/test-lib.sh +4 −2

diff --git a/t/test-lib.sh b/t/test-lib.sh
index bd4b02e..51e4a88 100644
--- a/t/test-lib.sh
+++ b/t/test-lib.sh
@@ -907,9 +907,11 @@ yes () {
 		y="$*"
 	fi
 
-	while echo "$y"
+	i=0
+	while test $i -lt 99
 	do
-		:
+		echo "$y"
+		i=$(($i+1))
 	done
 }

-- 
2.7.0.118.g90056ae

Johannes Schindelin· Feb 3, 2016, 12:11 UTC · re: Johannes Sixt · lore

Re: [PATCH v2] test-lib: limit the output of the yes utility

Hi Hannes,

On Tue, 2 Feb 2016, Johannes Sixt wrote:

> I'm not 100% satisfied with your version because it stomps on
> short-and-sweet shell variables $i and $y.

Please note that $y was already stomped on before my patch.

> But then the utility only occurs upstream of a pipeline in a separate
> process, so that does not matter.

That was my thinking, too.

> Please allow me to pass authorship to you, since the patch text is
> now all yours, and to forge your sign-off.

Fair enough, Dscho

patchtest-lib: limit the output of the yes utility