qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Andrey Shinkevich <andrey.shinkevich@virtuozzo.com>
To: John Snow <jsnow@redhat.com>,
	"qemu-devel@nongnu.org" <qemu-devel@nongnu.org>,
	"qemu-block@nongnu.org" <qemu-block@nongnu.org>
Cc: "kwolf@redhat.com" <kwolf@redhat.com>,
	Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>,
	Denis Lunev <den@virtuozzo.com>,
	"mreitz@redhat.com" <mreitz@redhat.com>
Subject: Re: [Qemu-devel] [Qemu-block] [PATCH v5 1/6] iotests: allow Valgrind checking all QEMU processes
Date: Sun, 25 Aug 2019 15:26:33 +0000	[thread overview]
Message-ID: <c1a0bac0-8628-87d6-8564-68287bebd705@virtuozzo.com> (raw)
In-Reply-To: <e1448863-8133-dec0-bce3-7b13664deb6b@redhat.com>



On 16/08/2019 01:49, John Snow wrote:
> 
> 
> On 7/19/19 12:30 PM, Andrey Shinkevich wrote:
>> With the '-valgrind' option, let all the QEMU processes be run under
>> the Valgrind tool. The Valgrind own parameters may be set with its
>> environment variable VALGRIND_OPTS, e.g.
>> VALGRIND_OPTS="--leak-check=yes" ./check -qcow2 -valgrind <test#>
>> or they may be listed in the Valgrind checked file ./.valgrindrc or
>> ~/.valgrindrc like
>> --memcheck:leak-check=no
>> --memcheck:track-origins=yes
>> When QEMU-IO process is being killed, the shell report refers to the
>> text of the command in _qemu_io_wrapper(), which was modified with this
>> patch. So, the benchmark output for the tests 039, 061 and 137 is to be
>> changed also.
>>
> 
> Oh, weird. "VALGRIND_QEMU=y" actually has just meant ... valgrind
> qemu-io. OK.
> 
>> Signed-off-by: Andrey Shinkevich <andrey.shinkevich@virtuozzo.com>
>> ---
>>   tests/qemu-iotests/039.out   | 30 ++++---------------
>>   tests/qemu-iotests/061.out   | 12 ++------
>>   tests/qemu-iotests/137.out   |  6 +---
>>   tests/qemu-iotests/common.rc | 69 ++++++++++++++++++++++++++++++++------------
>>   4 files changed, 59 insertions(+), 58 deletions(-)
>>
>> diff --git a/tests/qemu-iotests/039.out b/tests/qemu-iotests/039.out
>> index 724d7b2..972c6c0 100644
>> --- a/tests/qemu-iotests/039.out
>> +++ b/tests/qemu-iotests/039.out
>> @@ -11,11 +11,7 @@ No errors were found on the image.
>>   Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728
>>   wrote 512/512 bytes at offset 0
>>   512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
>> -./common.rc: Killed                  ( if [ "${VALGRIND_QEMU}" == "y" ]; then
>> -    exec valgrind --log-file="${VALGRIND_LOGFILE}" --error-exitcode=99 "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@";
>> -else
>> -    exec "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@";
>> -fi )
>> +./common.rc: Killed                  ( _qemu_proc_wrapper "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" )
>>   incompatible_features     0x1
>>   ERROR cluster 5 refcount=0 reference=1
>>   ERROR OFLAG_COPIED data cluster: l2_entry=8000000000050000 refcount=0
>> @@ -50,11 +46,7 @@ read 512/512 bytes at offset 0
>>   Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728
>>   wrote 512/512 bytes at offset 0
>>   512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
>> -./common.rc: Killed                  ( if [ "${VALGRIND_QEMU}" == "y" ]; then
>> -    exec valgrind --log-file="${VALGRIND_LOGFILE}" --error-exitcode=99 "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@";
>> -else
>> -    exec "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@";
>> -fi )
>> +./common.rc: Killed                  ( _qemu_proc_wrapper "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" )
>>   incompatible_features     0x1
>>   ERROR cluster 5 refcount=0 reference=1
>>   Rebuilding refcount structure
>> @@ -68,11 +60,7 @@ incompatible_features     0x0
>>   Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728
>>   wrote 512/512 bytes at offset 0
>>   512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
>> -./common.rc: Killed                  ( if [ "${VALGRIND_QEMU}" == "y" ]; then
>> -    exec valgrind --log-file="${VALGRIND_LOGFILE}" --error-exitcode=99 "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@";
>> -else
>> -    exec "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@";
>> -fi )
>> +./common.rc: Killed                  ( _qemu_proc_wrapper "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" )
>>   incompatible_features     0x0
>>   No errors were found on the image.
>>   
>> @@ -91,11 +79,7 @@ No errors were found on the image.
>>   Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728
>>   wrote 512/512 bytes at offset 0
>>   512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
>> -./common.rc: Killed                  ( if [ "${VALGRIND_QEMU}" == "y" ]; then
>> -    exec valgrind --log-file="${VALGRIND_LOGFILE}" --error-exitcode=99 "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@";
>> -else
>> -    exec "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@";
>> -fi )
>> +./common.rc: Killed                  ( _qemu_proc_wrapper "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" )
>>   incompatible_features     0x1
>>   ERROR cluster 5 refcount=0 reference=1
>>   ERROR OFLAG_COPIED data cluster: l2_entry=8000000000050000 refcount=0
>> @@ -105,11 +89,7 @@ Data may be corrupted, or further writes to the image may corrupt it.
>>   Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728
>>   wrote 512/512 bytes at offset 0
>>   512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
>> -./common.rc: Killed                  ( if [ "${VALGRIND_QEMU}" == "y" ]; then
>> -    exec valgrind --log-file="${VALGRIND_LOGFILE}" --error-exitcode=99 "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@";
>> -else
>> -    exec "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@";
>> -fi )
>> +./common.rc: Killed                  ( _qemu_proc_wrapper "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" )
>>   incompatible_features     0x0
>>   No errors were found on the image.
>>   *** done
>> diff --git a/tests/qemu-iotests/061.out b/tests/qemu-iotests/061.out
>> index 1aa7d37..8cb57eb 100644
>> --- a/tests/qemu-iotests/061.out
>> +++ b/tests/qemu-iotests/061.out
>> @@ -118,11 +118,7 @@ No errors were found on the image.
>>   Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864
>>   wrote 131072/131072 bytes at offset 0
>>   128 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
>> -./common.rc: Killed                  ( if [ "${VALGRIND_QEMU}" == "y" ]; then
>> -    exec valgrind --log-file="${VALGRIND_LOGFILE}" --error-exitcode=99 "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@";
>> -else
>> -    exec "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@";
>> -fi )
>> +./common.rc: Killed                  ( _qemu_proc_wrapper "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" )
>>   magic                     0x514649fb
>>   version                   3
>>   backing_file_offset       0x0
>> @@ -280,11 +276,7 @@ No errors were found on the image.
>>   Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864
>>   wrote 131072/131072 bytes at offset 0
>>   128 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
>> -./common.rc: Killed                  ( if [ "${VALGRIND_QEMU}" == "y" ]; then
>> -    exec valgrind --log-file="${VALGRIND_LOGFILE}" --error-exitcode=99 "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@";
>> -else
>> -    exec "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@";
>> -fi )
>> +./common.rc: Killed                  ( _qemu_proc_wrapper "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" )
>>   magic                     0x514649fb
>>   version                   3
>>   backing_file_offset       0x0
>> diff --git a/tests/qemu-iotests/137.out b/tests/qemu-iotests/137.out
>> index 22d59df..7fed5e6 100644
>> --- a/tests/qemu-iotests/137.out
>> +++ b/tests/qemu-iotests/137.out
>> @@ -35,11 +35,7 @@ Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864
>>   qemu-io: Unsupported value 'blubb' for qcow2 option 'overlap-check'. Allowed are any of the following: none, constant, cached, all
>>   wrote 512/512 bytes at offset 0
>>   512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
>> -./common.rc: Killed                  ( if [ "${VALGRIND_QEMU}" == "y" ]; then
>> -    exec valgrind --log-file="${VALGRIND_LOGFILE}" --error-exitcode=99 "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@";
>> -else
>> -    exec "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@";
>> -fi )
>> +./common.rc: Killed                  ( _qemu_proc_wrapper "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" )
>>   incompatible_features     0x0
>>   Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864
>>   wrote 65536/65536 bytes at offset 0
>> diff --git a/tests/qemu-iotests/common.rc b/tests/qemu-iotests/common.rc
>> index 5502c3d..6e461a1 100644
>> --- a/tests/qemu-iotests/common.rc
>> +++ b/tests/qemu-iotests/common.rc
>> @@ -60,19 +60,52 @@ if ! . ./common.config
>>       exit 1
>>   fi
>>   
>> +_qemu_proc_wrapper()
>> +{
>> +    local VALGRIND_LOGFILE="$1"
>> +    shift
>> +    if [ "${VALGRIND_QEMU}" == "y" ]; then
>> +        exec valgrind --log-file="${VALGRIND_LOGFILE}" --error-exitcode=99 "$@"
>> +    else
>> +        exec "$@"
>> +    fi
>> +}
>> +
> 
> Why do we need a second wrapper? I get nervous with each new wrapper we
> make, because I feel like it has unintended consequences with pipe
> handling and so on in the test dispatcher scripts.
> 

The _qemu_proc_wrapper() is acctually the function rather than a new 
wrapper. The suffix _wrapper can be changed to mislead a reader not. The 
routine code originates from the _qemu_io_wrapper() to repeat the same 
functionality for the QEMU processes other than qemu-io and to keep the 
uniformity. It is also true for the function _qemu_proc_valgrind_log() 
that dumps Valgrind reports in case of errors. The behavior is the same 
to one in the _qemu_io_wrapper(). Actually, nothing new except that 
_qemu_nbd_wrapper() gets the subshell exec to keep the uniformity.

Andrey

>> +_qemu_proc_valgrind_log()
>> +{
>> +    local VALGRIND_LOGFILE="$1"
>> +    local RETVAL="$2"
>> +    if [ "${VALGRIND_QEMU}" == "y" ]; then
>> +        if [ $RETVAL == 99 ]; then
>> +            cat "${VALGRIND_LOGFILE}"
>> +        fi
>> +        rm -f "${VALGRIND_LOGFILE}"
>> +    fi
>> +}
>> +
>>   _qemu_wrapper()
>>   {
>> +    local VALGRIND_LOGFILE="${TEST_DIR}"/$$.valgrind
>>       (
>>           if [ -n "${QEMU_NEED_PID}" ]; then
>>               echo $BASHPID > "${QEMU_TEST_DIR}/qemu-${_QEMU_HANDLE}.pid"
>>           fi
>> -        exec "$QEMU_PROG" $QEMU_OPTIONS "$@"
>> +        _qemu_proc_wrapper "${VALGRIND_LOGFILE}" "$QEMU_PROG" $QEMU_OPTIONS "$@"
> 
> Can we not inline that logic here? especially because:
> 
>>       )
>> +    RETVAL=$?
>> +    _qemu_proc_valgrind_log "${VALGRIND_LOGFILE}" $RETVAL
> 
> We're already making valgrind calls here anyway.
> 

We can of cause, but we would repeat that for all other 
_qemu_*_wrapper() functions as well. The _qemu_proc_wrapper() runs a 
process under the Valgrind if VALGRIND_QEMU=y and the 
_qemu_proc_valgrind_log() dumps Valgrind error reports, if any and 
removes the Valgrind temporary log file. Again, the same pattern is used 
as it is in the current implementation of _qemu_io_wrapper(). I have 
just copied the functionality to all other QEMU wrapper functions.

Andrey

>> +    return $RETVAL
>>   }
>>   
>>   _qemu_img_wrapper()
>>   {
>> -    (exec "$QEMU_IMG_PROG" $QEMU_IMG_OPTIONS "$@")
>> +    local VALGRIND_LOGFILE="${TEST_DIR}"/$$.valgrind
>> +    (
>> +        _qemu_proc_wrapper "${VALGRIND_LOGFILE}" "$QEMU_IMG_PROG" $QEMU_IMG_OPTIONS "$@"
>> +    )
>> +    RETVAL=$?
>> +    _qemu_proc_valgrind_log "${VALGRIND_LOGFILE}" $RETVAL
>> +    return $RETVAL
>>   }
>>   
>>   _qemu_io_wrapper()
>> @@ -85,36 +118,36 @@ _qemu_io_wrapper()
>>               QEMU_IO_ARGS="--object secret,id=keysec0,data=$IMGKEYSECRET $QEMU_IO_ARGS"
>>           fi
>>       fi
>> -    local RETVAL
>>       (
>> -        if [ "${VALGRIND_QEMU}" == "y" ]; then
>> -            exec valgrind --log-file="${VALGRIND_LOGFILE}" --error-exitcode=99 "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@"
>> -        else
>> -            exec "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@"
>> -        fi
>> +        _qemu_proc_wrapper "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@"
>>       )
>>       RETVAL=$?
>> -    if [ "${VALGRIND_QEMU}" == "y" ]; then
>> -        if [ $RETVAL == 99 ]; then
>> -            cat "${VALGRIND_LOGFILE}"
>> -        fi
>> -        rm -f "${VALGRIND_LOGFILE}"
>> -    fi
>> -    (exit $RETVAL)
>> +    _qemu_proc_valgrind_log "${VALGRIND_LOGFILE}" $RETVAL
>> +    return $RETVAL
>>   }
>>   
>>   _qemu_nbd_wrapper()
>>   {
>> -    "$QEMU_NBD_PROG" --pid-file="${QEMU_TEST_DIR}/qemu-nbd.pid" \
>> -                     $QEMU_NBD_OPTIONS "$@"
>> +    local VALGRIND_LOGFILE="${TEST_DIR}"/$$.valgrind
>> +    (
>> +        _qemu_proc_wrapper "${VALGRIND_LOGFILE}" "$QEMU_NBD_PROG" \
>> +            --pid-file="${QEMU_TEST_DIR}/qemu-nbd.pid" $QEMU_NBD_OPTIONS "$@"
>> +    )
>> +    RETVAL=$?
>> +    _qemu_proc_valgrind_log "${VALGRIND_LOGFILE}" $RETVAL
>> +    return $RETVAL
>>   }
>>   
>>   _qemu_vxhs_wrapper()
>>   {
>> +    local VALGRIND_LOGFILE="${TEST_DIR}"/$$.valgrind
>>       (
>>           echo $BASHPID > "${TEST_DIR}/qemu-vxhs.pid"
>> -        exec "$QEMU_VXHS_PROG" $QEMU_VXHS_OPTIONS "$@"
>> +        _qemu_proc_wrapper "${VALGRIND_LOGFILE}" "$QEMU_VXHS_PROG" $QEMU_VXHS_OPTIONS "$@"
>>       )
>> +    RETVAL=$?
>> +    _qemu_proc_valgrind_log "${VALGRIND_LOGFILE}" $RETVAL
>> +    return $RETVAL
>>   }
>>   
>>   export QEMU=_qemu_wrapper
>>
> 
> Only other thought: at the moment, valgrind turns on valgrind for
> qemu-io; this wraps many more tools and potentially slows down iotests a
> lot. If there are problems, we might want easy access to turn /some/ but
> not all of these options off again.
> 
> 
> do we want to be able to specify which subprocesses we use valgrind on,
> Perhaps with environment variables for fine-tuning?
> 
> QEMU_VALGRIND_QEMU_IO = "off"
> QEMU_VALGRIND_QEMU = "off"
> QEMU_VALGRIND_NBD = "off"
> 
> (Only an idle question, not necessary for this series IMO.)
> 

-- 
With the best regards,
Andrey Shinkevich

  reply	other threads:[~2019-08-25 15:27 UTC|newest]

Thread overview: 29+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-07-19 16:30 [Qemu-devel] [PATCH v5 0/6] Allow Valgrind checking all QEMU processes Andrey Shinkevich
2019-07-19 16:30 ` [Qemu-devel] [PATCH v5 1/6] iotests: allow " Andrey Shinkevich
2019-08-15 22:49   ` [Qemu-devel] [Qemu-block] " John Snow
2019-08-25 15:26     ` Andrey Shinkevich [this message]
2019-08-27 19:56       ` John Snow
2019-08-28 15:04         ` Andrey Shinkevich
2019-07-19 16:30 ` [Qemu-devel] [PATCH v5 2/6] iotests: exclude killed processes from running under Valgrind Andrey Shinkevich
2019-08-16  0:40   ` [Qemu-devel] [Qemu-block] " John Snow
2019-07-19 16:30 ` [Qemu-devel] [PATCH v5 3/6] iotests: Add casenotrun report to bash tests Andrey Shinkevich
2019-08-16  0:44   ` [Qemu-devel] [Qemu-block] " John Snow
2019-08-16 20:33     ` Cleber Rosa
2019-08-25 13:03       ` Andrey Shinkevich
2019-07-19 16:30 ` [Qemu-devel] [PATCH v5 4/6] iotests: Valgrind fails with nonexistent directory Andrey Shinkevich
2019-08-16  0:55   ` [Qemu-devel] [Qemu-block] " John Snow
2019-08-25 15:24     ` Andrey Shinkevich
2019-08-27 19:45       ` John Snow
2019-08-28 15:12         ` Andrey Shinkevich
2019-07-19 16:30 ` [Qemu-devel] [PATCH v5 5/6] iotests: extended timeout under Valgrind Andrey Shinkevich
2019-08-16  0:58   ` [Qemu-devel] [Qemu-block] " John Snow
2019-07-19 16:30 ` [Qemu-devel] [PATCH v5 6/6] iotests: extend sleeping time " Andrey Shinkevich
2019-08-16  1:01   ` [Qemu-devel] [Qemu-block] " John Snow
2019-08-23 15:27     ` Vladimir Sementsov-Ogievskiy
2019-08-27 19:42       ` John Snow
2019-08-28 15:24         ` Andrey Shinkevich
2019-08-28 17:27           ` John Snow
2019-08-25 10:13     ` Andrey Shinkevich
2019-08-06 16:24 ` [Qemu-devel] [PATCH v5 0/6] Allow Valgrind checking all QEMU processes Andrey Shinkevich
2019-08-16 20:05 ` Cleber Rosa
2019-08-25 10:30   ` Andrey Shinkevich

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=c1a0bac0-8628-87d6-8564-68287bebd705@virtuozzo.com \
    --to=andrey.shinkevich@virtuozzo.com \
    --cc=den@virtuozzo.com \
    --cc=jsnow@redhat.com \
    --cc=kwolf@redhat.com \
    --cc=mreitz@redhat.com \
    --cc=qemu-block@nongnu.org \
    --cc=qemu-devel@nongnu.org \
    --cc=vsementsov@virtuozzo.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).