Re: [PATCH 1/2] iotests: Make _filter_img_create more active

[Maxim Levitsky]

On Wed, 2020-06-17 at 15:50 +0200, Max Reitz wrote:
> On 17.06.20 13:46, Maxim Levitsky wrote:
> > On Tue, 2020-06-16 at 15:17 +0200, Max Reitz wrote:
> > > Right now, _filter_img_create just filters out everything that looks
> > > format-dependent, and applies some filename filters.  That means that we
> > > have to add another filter line every time some format gets a new
> > > creation option.  This can be avoided by instead discarding everything
> > > and just keeping what we know is format-independent (format, size,
> > > backing file, encryption information[1], preallocation) or just
> > > interesting to have in the reference output (external data file path).
> > > 
> > > Furthermore, we probably want to sort these options.  Format drivers are
> > > not required to define them in any specific order, so the output is
> > > effectively random (although this has never bothered us until now).  We
> > > need a specific order for our reference outputs, though.  Unfortunately,
> > > just using a plain "sort" would change a lot of existing reference
> > > outputs, so we have to pre-filter the option keys to keep our existing
> > > order (fmt, size, backing*, data, encryption info, preallocation).
> > > 
> > > [1] Actually, the only thing that is really important is whether
> > >     encryption is enabled or not.  A patch by Maxim thus removes all
> > >     other "encrypt.*" options from the output:
> > >     https://lists.nongnu.org/archive/html/qemu-block/2020-06/msg00339.html
> > >     But that patch needs to come later so we can get away with changing
> > >     as few reference outputs in this patch here as possible.
> > > 
> > > Signed-off-by: Max Reitz <mreitz@redhat.com>
> > > ---
> > >  tests/qemu-iotests/112.out       |   2 +-
> > >  tests/qemu-iotests/153           |   9 ++-
> > >  tests/qemu-iotests/common.filter | 100 +++++++++++++++++++++++--------
> > >  3 files changed, 81 insertions(+), 30 deletions(-)
> > > 
> > > diff --git a/tests/qemu-iotests/112.out b/tests/qemu-iotests/112.out
> > > index ae0318cabe..182655dbf6 100644
> > > --- a/tests/qemu-iotests/112.out
> > > +++ b/tests/qemu-iotests/112.out
> > > @@ -5,7 +5,7 @@ QA output created by 112
> > >  qemu-img: TEST_DIR/t.IMGFMT: Refcount width must be a power of two and 
> > > may not exceed 64 bits
> > >  Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864
> > >  qemu-img: TEST_DIR/t.IMGFMT: Refcount width must be a power of two and 
> > > may not exceed 64 bits
> > > -Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 refcount_bits=-1
> > > +Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864
> > >  qemu-img: TEST_DIR/t.IMGFMT: Refcount width must be a power of two and 
> > > may not exceed 64 bits
> > >  Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864
> > >  qemu-img: TEST_DIR/t.IMGFMT: Refcount width must be a power of two and 
> > > may not exceed 64 bits
> > > diff --git a/tests/qemu-iotests/153 b/tests/qemu-iotests/153
> > > index cf961d3609..11e3d28841 100755
> > > --- a/tests/qemu-iotests/153
> > > +++ b/tests/qemu-iotests/153
> > > @@ -167,11 +167,10 @@ done
> > >  
> > >  echo
> > >  echo "== Creating ${TEST_IMG}.[abc] ==" | _filter_testdir
> > > -(
> > > -    $QEMU_IMG create -f qcow2 "${TEST_IMG}.a" -b "${TEST_IMG}"
> > > -    $QEMU_IMG create -f qcow2 "${TEST_IMG}.b" -b "${TEST_IMG}"
> > > -    $QEMU_IMG create -f qcow2 "${TEST_IMG}.c" -b "${TEST_IMG}.b"
> > > -) | _filter_img_create
> > > +$QEMU_IMG create -f qcow2 "${TEST_IMG}.a" -b "${TEST_IMG}" | 
> > > _filter_img_create
> > > +$QEMU_IMG create -f qcow2 "${TEST_IMG}.b" -b "${TEST_IMG}" | 
> > > _filter_img_create
> > > +$QEMU_IMG create -f qcow2 "${TEST_IMG}.c" -b "${TEST_IMG}.b" \
> > > +    | _filter_img_create
> > >  
> > >  echo
> > >  echo "== Two devices sharing the same file in backing chain =="
> > 
> > I guess this is done because now the filter expectes only a single
> > qemu-img output.
> 
> Yes, that’s right.
> 
> > IMHO this is better anyway.
> > 
> > > diff --git a/tests/qemu-iotests/common.filter 
> > > b/tests/qemu-iotests/common.filter
> > > index 03e4f71808..f104ad7a9b 100644
> > > --- a/tests/qemu-iotests/common.filter
> > > +++ b/tests/qemu-iotests/common.filter
> > > @@ -122,38 +122,90 @@ _filter_actual_image_size()
> > >  # replace driver-specific options in the "Formatting..." line
> > >  _filter_img_create()
> > >  {
> > > -    data_file_filter=()
> > > -    if data_file=$(_get_data_file "$TEST_IMG"); then
> > > -        data_file_filter=(-e "s# data_file=$data_file##")
> > > +    # Keep QMP output unchanged
> > > +    qmp_pre=''
> > > +    qmp_post=''
> > > +    to_filter=''
> > > +
> > > +    while read -r line; do
> > > +        if echo "$line" | grep -q '^{.*}$'; then
> > > +            if [ -z "$to_filter" ]; then
> > > +                # Use $'\n' so the newline is not dropped on variable
> > > +                # expansion
> > > +                qmp_pre="$qmp_pre$line"$'\n'
> > > +            else
> > > +                qmp_post="$qmp_post$line"$'\n'
> > > +            fi
> > > +        else
> > > +            to_filter="$to_filter$line"$'\n'
> > > +        fi
> > > +    done
> > 
> > The above code basically assumes that qmp output starts with '{' and ends 
> > with '}'
> > which I guess is fair, and then it assumes that we can have set of qmp 
> > commands prior
> > to qemu-img line and another set of qmp commands after it.
> > To me it feels like we should have another filter for that, since
> > qemu-img itself doesn't use qmp.
> 
> Yes, but drive-backup and drive-mirror with mode=absolute-paths use
> bdrv_img_create() (quiet=false), which emits the “Formatting” line, too.
>  So some QMP commands may emit it and then require the same filtering as
> qemu-img create.

Ah. Do we need this though? 
Assuming yes, maybe we should create a new filter called something like 
_filter_drive_backup_mirror
or something like that that would filter qmp output, and pipe the 'Formatting' 
line
to _filter_img_create which then can have the qmp parsing part removed?

> 
> Not having to do that would certainly make things simpler.
> 
> > Which test needs it?
> 
> 141.
> 
> > > +
> > > +    readarray -td '' formatting_line < \
> > > +        <(echo "$to_filter" | sed -e 's/, fmt=/\x0/')
> > OK, took me a while to understand what this does, but looks OK.
> 
> Maybe I should at least put a comment like “Split $to_filter into the
> pre-options part (before ", fmt=") and the options part (starting with
> "fmt=").
> 
> > > +
> > > +    filename_part=${formatting_line[0]}
> > > +    if [ -n "${formatting_line[1]}" ]; then
> > > +        options="fmt=${formatting_line[1]}"
> > > +    else
> > > +        options=''
> > >      fi
> > > 
> > OK.
> > 
> > >  
> > > -    $SED "${data_file_filter[@]}" \
> > > +    # Set grep_data_file to '\|data_file' to keep it; make it empty
> > > +    # to drop it.
> > > +    # We want to drop it if it is part of the global $IMGOPTS, and we
> > > +    # want to keep it otherwise (if the test specifically wants to
> > > +    # test data files).
> > > +    grep_data_file='\|data_file'
> > > +    if _get_data_file "$TEST_IMG" > /dev/null; then
> > > +        grep_data_file=''
> > > +    fi
> > > +
> > > +    filename_filters=(
> > >          -e "s#$REMOTE_TEST_DIR#TEST_DIR#g" \
> > >          -e "s#$IMGPROTO:$TEST_DIR#TEST_DIR#g" \
> > >          -e "s#$TEST_DIR#TEST_DIR#g" \
> > >          -e "s#$SOCK_DIR#SOCK_DIR#g" \
> > >          -e "s#$IMGFMT#IMGFMT#g" \
> > >          -e 's#nbd:127.0.0.1:[0-9]\\+#TEST_DIR/t.IMGFMT#g' \
> > > -        -e 's#nbd+unix:///\??socket=SOCK_DIR/nbd#TEST_DIR/t.IMGFMT#g' \
> > > -        -e "s# encryption=off##g" \
> > > -        -e "s# cluster_size=[0-9]\\+##g" \
> > > -        -e "s# table_size=[0-9]\\+##g" \
> > > -        -e "s# compat=[^ ]*##g" \
> > > -        -e "s# compat6=\\(on\\|off\\)##g" \
> > > -        -e "s# static=\\(on\\|off\\)##g" \
> > > -        -e "s# zeroed_grain=\\(on\\|off\\)##g" \
> > > -        -e "s# subformat=[^ ]*##g" \
> > > -        -e "s# adapter_type=[^ ]*##g" \
> > > -        -e "s# hwversion=[^ ]*##g" \
> > > -        -e "s# lazy_refcounts=\\(on\\|off\\)##g" \
> > > -        -e "s# block_size=[0-9]\\+##g" \
> > > -        -e "s# block_state_zero=\\(on\\|off\\)##g" \
> > > -        -e "s# log_size=[0-9]\\+##g" \
> > > -        -e "s# refcount_bits=[0-9]\\+##g" \
> > > -        -e "s# key-secret=[a-zA-Z0-9]\\+##g" \
> > > -        -e "s# iter-time=[0-9]\\+##g" \
> > > -        -e "s# force_size=\\(on\\|off\\)##g" \
> > > -        -e "s# compression_type=[a-zA-Z0-9]\\+##g"
> > > +        -e 's#nbd+unix:///\??socket=SOCK_DIR/nbd#TEST_DIR/t.IMGFMT#g'
> > > +    )
> > > +
> > > +    filename_part=$(echo "$filename_part" | $SED 
> > > "${filename_filters[@]}")
> > Looks good as well.
> > 
> > > +
> > > +    # Break the option line before each option (preserving pre-existing
> > > +    # line breaks by replacing them by \0 and restoring them at the end),
> > > +    # then filter out the options we want to keep and sort them according
> > > +    # to some order that all block drivers used at the time of writing
> > > +    # this function.
> > > +    options=$(
> > > +        echo "$options" \
> > > +        | tr '\n' '\0' \
> > > +        | sed -e 's/\x0$//' -e 's/ \([a-z0-9_.-]*\)=/\n\1=/g' \
> > You sometimes use $SED and sometimes sed. Is this intentional?
> 
> Oops, no.  It should be $SED everywhere.
> 
> > > +        | grep -ae 
> > > "^\(fmt\\|size\\|backing\\|preallocation\\|encrypt$grep_data_file\\)" \
> > > +        | $SED "${filename_filters[@]}" \
> > > +            -e 's/^\(fmt\)/0-\1/' \
> > > +            -e 's/^\(size\)/1-\1/' \
> > > +            -e 's/^\(backing\)/2-\1/' \
> > > +            -e 's/^\(data_file\)/3-\1/' \
> > > +            -e 's/^\(encryption\)/4-\1/' \
> > > +            -e 's/^\(encrypt\.format\)/5-\1/' \
> > > +            -e 's/^\(encrypt\.key-secret\)/6-\1/' \
> > > +            -e 's/^\(encrypt\.iter-time\)/7-\1/' \
> > > +            -e 's/^\(preallocation\)/8-\1/' \
> > All right, I understand this now, but do we have to do this?
> > Maybe it is better to just update the outputs once to avoid keeping
> > the custom sort order?
> 
> Well.  I don’t know.  The advantage of doing this is that I can’t
> accidentally miss updating any reference outputs that aren’t used on my
> machine (like 026.out.nocache or 051.out).

Fair enough. A follow up patch can always be made to fix this.

> 
> So we don’t strictly have to do this, but I found this to be simpler.
> If someone wants to drop this and in turn update all reference outputs,
> they’ll be my guest, but I preferred not to do that myself.
> 
> > > +        | sort \
> > > +        | $SED -e 's/^[0-9]-//' \
> > > +        | tr '\n\0' ' \n' \
> > > +        | $SED -e 's/^ *$//' -e 's/ *$//'
> > > +    )
> > 
> > For the above bash pipeline overall: It was hard to decipher :-), but I 
> > must admit
> > I learned something from it.
> 
> I definitely learned something while working on this, too.  I’m not sure
> whether I also gained any useful knowledge, though. O:)

In my past experience, any knowelege eventually turns out to be useful.

> 
> > > +
> > > +    echo -n "$qmp_pre"
> > > +    if [ -n "$options" ]; then
> > > +        echo "$filename_part, $options"
> > > +    elif [ -n "$filename_part" ]; then
> > > +        echo "$filename_part"
> > > +    fi
> > > +    echo -n "$qmp_post"
> > >  }
> > >  
> > >  _filter_img_create_size()
> > 
> > Overall I like the idea of this patch.
> 
> Good!
> 
> Thanks for reviewing! :)


Best regards,
        Maxim Levitsky