Message ID | 20200227072953.25445-5-dplotnikov@virtuozzo.com (mailing list archive) |
---|---|
State | New, archived |
Headers | show |
Series | qcow2: Implement zstd cluster compression method | expand |
27.02.2020 10:29, Denis Plotnikov wrote: > After adding compression type feature to qcow2 format, qemu framework > commands reporting the image settingd, e.g. "qemu-img create", started > reporting the compression type for the image which breaks the iotests > output matching. > > To fix it, add compression_type=zlib to the list of filtered image parameters. So, the first patch breaks iotests? Than it should be merged into first patch or moved before it, to not break git bisect. > > Signed-off-by: Denis Plotnikov <dplotnikov@virtuozzo.com> > --- > tests/qemu-iotests/common.filter | 3 ++- > 1 file changed, 2 insertions(+), 1 deletion(-) > > diff --git a/tests/qemu-iotests/common.filter b/tests/qemu-iotests/common.filter > index 3f8ee3e5f7..c6962d199c 100644 > --- a/tests/qemu-iotests/common.filter > +++ b/tests/qemu-iotests/common.filter > @@ -152,7 +152,8 @@ _filter_img_create() > -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# force_size=\\(on\\|off\\)##g" \ > + -e "s# compression_type=zlib##g" > } > > _filter_img_info() >
On 2/27/20 1:29 AM, Denis Plotnikov wrote: > After adding compression type feature to qcow2 format, qemu framework > commands reporting the image settingd, e.g. "qemu-img create", started settings > reporting the compression type for the image which breaks the iotests > output matching. > > To fix it, add compression_type=zlib to the list of filtered image parameters. > > Signed-off-by: Denis Plotnikov <dplotnikov@virtuozzo.com> > --- > tests/qemu-iotests/common.filter | 3 ++- > 1 file changed, 2 insertions(+), 1 deletion(-) This should be squashed in to the patch that caused the breakage (3/8, if I'm right). > > diff --git a/tests/qemu-iotests/common.filter b/tests/qemu-iotests/common.filter > index 3f8ee3e5f7..c6962d199c 100644 > --- a/tests/qemu-iotests/common.filter > +++ b/tests/qemu-iotests/common.filter > @@ -152,7 +152,8 @@ _filter_img_create() > -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# force_size=\\(on\\|off\\)##g" \ > + -e "s# compression_type=zlib##g" Do you really want to hard-code just zlib, or should this be more generic as compression_type=[a-zA-Z0-9]\\+ as done on other lines like key-secret?
On 27.02.2020 17:03, Eric Blake wrote: > On 2/27/20 1:29 AM, Denis Plotnikov wrote: >> After adding compression type feature to qcow2 format, qemu framework >> commands reporting the image settingd, e.g. "qemu-img create", started > > settings > >> reporting the compression type for the image which breaks the iotests >> output matching. >> >> To fix it, add compression_type=zlib to the list of filtered image >> parameters. >> >> Signed-off-by: Denis Plotnikov <dplotnikov@virtuozzo.com> >> --- >> tests/qemu-iotests/common.filter | 3 ++- >> 1 file changed, 2 insertions(+), 1 deletion(-) > > This should be squashed in to the patch that caused the breakage (3/8, > if I'm right). > >> >> diff --git a/tests/qemu-iotests/common.filter >> b/tests/qemu-iotests/common.filter >> index 3f8ee3e5f7..c6962d199c 100644 >> --- a/tests/qemu-iotests/common.filter >> +++ b/tests/qemu-iotests/common.filter >> @@ -152,7 +152,8 @@ _filter_img_create() >> -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# force_size=\\(on\\|off\\)##g" \ >> + -e "s# compression_type=zlib##g" > > Do you really want to hard-code just zlib, or should this be more > generic as compression_type=[a-zA-Z0-9]\\+ as done on other lines like > key-secret? When I did this I meant additional implicit check that the default compression type is zlib. But non of the other items in the filter don't do it. So I'll change it to be consistent. Thanks! Denis > >
diff --git a/tests/qemu-iotests/common.filter b/tests/qemu-iotests/common.filter index 3f8ee3e5f7..c6962d199c 100644 --- a/tests/qemu-iotests/common.filter +++ b/tests/qemu-iotests/common.filter @@ -152,7 +152,8 @@ _filter_img_create() -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# force_size=\\(on\\|off\\)##g" \ + -e "s# compression_type=zlib##g" } _filter_img_info()
After adding compression type feature to qcow2 format, qemu framework commands reporting the image settingd, e.g. "qemu-img create", started reporting the compression type for the image which breaks the iotests output matching. To fix it, add compression_type=zlib to the list of filtered image parameters. Signed-off-by: Denis Plotnikov <dplotnikov@virtuozzo.com> --- tests/qemu-iotests/common.filter | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)