Since we've recently added a ton of tests, the list starts to be a bit
of a mess and creates unneeded conflicts.
Let's order it alphabetically.
Signed-off-by: Maxime Ripard <[email protected]>
diff --git a/drivers/gpu/drm/tests/Makefile b/drivers/gpu/drm/tests/Makefile
index 91b70f7d2769..2d9f49b62ecb 100644
--- a/drivers/gpu/drm/tests/Makefile
+++ b/drivers/gpu/drm/tests/Makefile
@@ -1,5 +1,13 @@
# SPDX-License-Identifier: GPL-2.0
-obj-$(CONFIG_DRM_KUNIT_TEST) += drm_format_helper_test.o drm_damage_helper_test.o \
- drm_cmdline_parser_test.o drm_rect_test.o drm_format_test.o drm_plane_helper_test.o \
- drm_dp_mst_helper_test.o drm_framebuffer_test.o drm_buddy_test.o drm_mm_test.o
+obj-$(CONFIG_DRM_KUNIT_TEST) += \
+ drm_buddy_test.o \
+ drm_cmdline_parser_test.o \
+ drm_damage_helper_test.o \
+ drm_dp_mst_helper_test.o \
+ drm_format_helper_test.o \
+ drm_format_test.o \
+ drm_framebuffer_test.o \
+ drm_mm_test.o \
+ drm_plane_helper_test.o \
+ drm_rect_test.o
--
b4 0.10.0-dev-65ba7
Den 29.08.2022 15.11, skrev Maxime Ripard:
> Since we've recently added a ton of tests, the list starts to be a bit
>
> of a mess and creates unneeded conflicts.
>
>
>
> Let's order it alphabetically.
>
>
>
> Signed-off-by: Maxime Ripard <[email protected]>
>
>
>
Something has gone wrong with this patchset, there are double line endings.
I looked at the patchwork version and it look all right there so I
figured it might have fixed up the patches, but it failed:
git apply -v --check
/home/pi/tinydrm.gud-gadget/workdirs/tv_norm_gadget/500003.patch
Checking patch drivers/gpu/drm/tests/Makefile...
error: while searching for:
# SPDX-License-Identifier: GPL-2.0?
?
obj-$(CONFIG_DRM_KUNIT_TEST) += drm_format_helper_test.o
drm_damage_helper_test.o \?
drm_cmdline_parser_test.o drm_rect_test.o drm_format_test.o
drm_plane_helper_test.o \?
drm_dp_mst_helper_test.o drm_framebuffer_test.o drm_buddy_test.o
drm_mm_test.o?
error: patch failed: drivers/gpu/drm/tests/Makefile:1
error: drivers/gpu/drm/tests/Makefile: patch does not apply
ERROR: Failed check apply patch
pi@build-server:~/tinydrm.gud-gadget$ file
workdirs/tv_norm_gadget/500003.patch
workdirs/tv_norm_gadget/500003.patch: unified diff output, ASCII text,
with CRLF, LF line terminators
Noralf.
> diff --git a/drivers/gpu/drm/tests/Makefile b/drivers/gpu/drm/tests/Makefile
>
> index 91b70f7d2769..2d9f49b62ecb 100644
>
> --- a/drivers/gpu/drm/tests/Makefile
>
> +++ b/drivers/gpu/drm/tests/Makefile
>
> @@ -1,5 +1,13 @@
>
> # SPDX-License-Identifier: GPL-2.0
>
>
>
> -obj-$(CONFIG_DRM_KUNIT_TEST) += drm_format_helper_test.o drm_damage_helper_test.o \
>
> - drm_cmdline_parser_test.o drm_rect_test.o drm_format_test.o drm_plane_helper_test.o \
>
> - drm_dp_mst_helper_test.o drm_framebuffer_test.o drm_buddy_test.o drm_mm_test.o
>
> +obj-$(CONFIG_DRM_KUNIT_TEST) += \
>
> + drm_buddy_test.o \
>
> + drm_cmdline_parser_test.o \
>
> + drm_damage_helper_test.o \
>
> + drm_dp_mst_helper_test.o \
>
> + drm_format_helper_test.o \
>
> + drm_format_test.o \
>
> + drm_framebuffer_test.o \
>
> + drm_mm_test.o \
>
> + drm_plane_helper_test.o \
>
> + drm_rect_test.o
>
>
>
On Mon, Aug 29, 2022 at 08:46:42PM +0200, Noralf Trønnes wrote:
> Something has gone wrong with this patchset, there are double line endings.
I noticed this, too, and I think the reason is because these patches were
generated with "b4 send -o", but actually sent using git-send-email. It's not
a use-case I've considered (or tested) and the breakage is because when b4
generates patches with "-o", they are written with CRLF line endings, which is
not something git-send-email expects.
Maxime, any reason you went this direction instead of just letting b4 send
these patches directly?
-K
Hi Konstantin,
On Mon, Aug 29, 2022 at 03:02:42PM -0400, Konstantin Ryabitsev wrote:
> On Mon, Aug 29, 2022 at 08:46:42PM +0200, Noralf Tr?nnes wrote:
> > Something has gone wrong with this patchset, there are double line endings.
>
> I noticed this, too, and I think the reason is because these patches were
> generated with "b4 send -o", but actually sent using git-send-email. It's not
> a use-case I've considered (or tested) and the breakage is because when b4
> generates patches with "-o", they are written with CRLF line endings, which is
> not something git-send-email expects.
You're right
> Maxime, any reason you went this direction instead of just letting b4 send
> these patches directly?
For some reason, b4 send doesn't work for my setup and I haven't had the
time to look into why and let you know.
Whenever I used b4 send (on this series for example) I get:
$ b4 send
Converted the branch to 41 patches
Populating the To: and Cc: fields with automatically collected addresses
Invoking get_maintainer.pl for To: addresses
Invoking get_maintainer.pl for Cc: addresses
Will send the following messages:
---
To: David Airlie <[email protected]>
Daniel Vetter <[email protected]>
Maarten Lankhorst <[email protected]>
Maxime Ripard <[email protected]>
Thomas Zimmermann <[email protected]>
Noralf Tr?nnes <[email protected]>
Emma Anholt <[email protected]>
Ben Skeggs <[email protected]>
Karol Herbst <[email protected]>
Lyude Paul <[email protected]>
Jani Nikula <[email protected]>
Joonas Lahtinen <[email protected]>
Rodrigo Vivi <[email protected]>
Tvrtko Ursulin <[email protected]>
Chen-Yu Tsai <[email protected]>
Jernej Skrabec <[email protected]>
Samuel Holland <[email protected]>
Philipp Zabel <[email protected]>
Cc: [email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
Dom Cobley <[email protected]>
Maxime Ripard <[email protected]>
Geert Uytterhoeven <[email protected]>
Hans de Goede <[email protected]>
Dave Stevenson <[email protected]>
Mateusz Kwiatkowski <[email protected]>
Phil Elwell <[email protected]>
---
[PATCH v2 00/41] drm: Analog TV Improvements
[PATCH v2 01/41] drm/tests: Order Kunit tests in Makefile
[PATCH v2 02/41] drm/tests: Add Kunit Helpers
[PATCH v2 03/41] drm/atomic-helper: Rename drm_atomic_helper_connector_tv_reset to avoid ambiguity
[PATCH v2 04/41] drm/connector: Rename subconnector state variable
[PATCH v2 05/41] drm/atomic: Add TV subconnector property to get/set_property
[PATCH v2 06/41] drm/connector: Rename legacy TV property
[PATCH v2 07/41] drm/connector: Only register TV mode property if present
[PATCH v2 08/41] drm/connector: Rename drm_mode_create_tv_properties
[PATCH v2 09/41] drm/connector: Add TV standard property
[PATCH v2 10/41] drm/modes: Add a function to generate analog display modes
[PATCH v2 11/41] drm/modes: Only consider bpp and refresh before options
[PATCH v2 12/41] drm/modes: parse_cmdline: Add support for named modes containing dashes
[PATCH v2 13/41] drm/client: Add some tests for drm_connector_pick_cmdline_mode()
[PATCH v2 14/41] drm/modes: Move named modes parsing to a separate function
[PATCH v2 15/41] drm/modes: Switch to named mode descriptors
[PATCH v2 16/41] drm/modes: Fill drm_cmdline mode from named modes
[PATCH v2 17/41] drm/connector: Add pixel clock to cmdline mode
[PATCH v2 18/41] drm/connector: Add a function to lookup a TV mode by its name
[PATCH v2 19/41] drm/modes: Introduce the tv_mode property as a command-line option
[PATCH v2 20/41] drm/modes: Properly generate a drm_display_mode from a named mode
[PATCH v2 21/41] drm/modes: Introduce more named modes
[PATCH v2 22/41] drm/atomic-helper: Add a TV properties reset helper
[PATCH v2 23/41] drm/atomic-helper: Add an analog TV atomic_check implementation
[PATCH v2 24/41] drm/vc4: vec: Remove empty mode_fixup
[PATCH v2 25/41] drm/vc4: vec: Convert to atomic helpers
[PATCH v2 26/41] drm/vc4: vec: Refactor VEC TV mode setting
[PATCH v2 27/41] drm/vc4: vec: Remove redundant atomic_mode_set
[PATCH v2 28/41] drm/vc4: vec: Fix timings for VEC modes
[PATCH v2 29/41] drm/vc4: vec: Switch for common modes
[PATCH v2 30/41] drm/vc4: vec: Fix definition of PAL-M mode
[PATCH v2 31/41] drm/vc4: vec: Use TV Reset implementation
[PATCH v2 32/41] drm/vc4: vec: Convert to the new TV mode property
[PATCH v2 33/41] drm/vc4: vec: Add support for more analog TV standards
[PATCH v2 34/41] drm/sun4i: tv: Remove unused mode_valid
[PATCH v2 35/41] drm/sun4i: tv: Convert to atomic hooks
[PATCH v2 36/41] drm/sun4i: tv: Merge mode_set into atomic_enable
[PATCH v2 37/41] drm/sun4i: tv: Remove useless function
[PATCH v2 38/41] drm/sun4i: tv: Remove useless destroy function
[PATCH v2 39/41] drm/sun4i: tv: Rename error label
[PATCH v2 40/41] drm/sun4i: tv: Add missing reset assertion
[PATCH v2 41/41] drm/sun4i: tv: Convert to the new TV mode property
---
Press Enter to send or Ctrl-C to abort
[PATCH v2 00/41] drm: Analog TV Improvements
[PATCH v2 01/41] drm/tests: Order Kunit tests in Makefile
[PATCH v2 02/41] drm/tests: Add Kunit Helpers
[PATCH v2 03/41] drm/atomic-helper: Rename drm_atomic_helper_connector_tv_reset to avoid ambiguity
[PATCH v2 04/41] drm/connector: Rename subconnector state variable
[PATCH v2 05/41] drm/atomic: Add TV subconnector property to get/set_property
[PATCH v2 06/41] drm/connector: Rename legacy TV property
[PATCH v2 07/41] drm/connector: Only register TV mode property if present
[PATCH v2 08/41] drm/connector: Rename drm_mode_create_tv_properties
[PATCH v2 09/41] drm/connector: Add TV standard property
[PATCH v2 10/41] drm/modes: Add a function to generate analog display modes
[PATCH v2 11/41] drm/modes: Only consider bpp and refresh before options
[PATCH v2 12/41] drm/modes: parse_cmdline: Add support for named modes containing dashes
[PATCH v2 13/41] drm/client: Add some tests for drm_connector_pick_cmdline_mode()
[PATCH v2 14/41] drm/modes: Move named modes parsing to a separate function
[PATCH v2 15/41] drm/modes: Switch to named mode descriptors
[PATCH v2 16/41] drm/modes: Fill drm_cmdline mode from named modes
[PATCH v2 17/41] drm/connector: Add pixel clock to cmdline mode
[PATCH v2 18/41] drm/connector: Add a function to lookup a TV mode by its name
[PATCH v2 19/41] drm/modes: Introduce the tv_mode property as a command-line option
[PATCH v2 20/41] drm/modes: Properly generate a drm_display_mode from a named mode
[PATCH v2 21/41] drm/modes: Introduce more named modes
[PATCH v2 22/41] drm/atomic-helper: Add a TV properties reset helper
[PATCH v2 23/41] drm/atomic-helper: Add an analog TV atomic_check implementation
[PATCH v2 24/41] drm/vc4: vec: Remove empty mode_fixup
[PATCH v2 25/41] drm/vc4: vec: Convert to atomic helpers
[PATCH v2 26/41] drm/vc4: vec: Refactor VEC TV mode setting
[PATCH v2 27/41] drm/vc4: vec: Remove redundant atomic_mode_set
[PATCH v2 28/41] drm/vc4: vec: Fix timings for VEC modes
[PATCH v2 29/41] drm/vc4: vec: Switch for common modes
[PATCH v2 30/41] drm/vc4: vec: Fix definition of PAL-M mode
[PATCH v2 31/41] drm/vc4: vec: Use TV Reset implementation
[PATCH v2 32/41] drm/vc4: vec: Convert to the new TV mode property
[PATCH v2 33/41] drm/vc4: vec: Add support for more analog TV standards
[PATCH v2 34/41] drm/sun4i: tv: Remove unused mode_valid
[PATCH v2 35/41] drm/sun4i: tv: Convert to atomic hooks
[PATCH v2 36/41] drm/sun4i: tv: Merge mode_set into atomic_enable
[PATCH v2 37/41] drm/sun4i: tv: Remove useless function
[PATCH v2 38/41] drm/sun4i: tv: Remove useless destroy function
[PATCH v2 39/41] drm/sun4i: tv: Rename error label
[PATCH v2 40/41] drm/sun4i: tv: Add missing reset assertion
[PATCH v2 41/41] drm/sun4i: tv: Convert to the new TV mode property
Connecting to /home/max/.local/bin/msmtpq:0
Failed to configure the smtp connection:
[Errno -2] Name or service not known
msmtpq is basically a sendmail-like tool to queue the mails if the
machine is offline. It is setup in my git config as
sendemail.smtpserver.
git-send-email works just fine, but somehow b4 send fails. I think it's
due to cmd_send calling b4.get_smtp() and expecting to get some
credentials out of it?
Maxime
(Dropping most cc's, who probably don't care about this thread.)
On Tue, Aug 30, 2022 at 10:30:27AM +0200, Maxime Ripard wrote:
> > I noticed this, too, and I think the reason is because these patches were
> > generated with "b4 send -o", but actually sent using git-send-email. It's not
> > a use-case I've considered (or tested) and the breakage is because when b4
> > generates patches with "-o", they are written with CRLF line endings, which is
> > not something git-send-email expects.
>
> You're right
OK, the latest master should allow you to do that without causing CRLF errors
like this series.
> For some reason, b4 send doesn't work for my setup and I haven't had the
> time to look into why and let you know.
Yeah, I see what the cause of it is:
> Connecting to /home/max/.local/bin/msmtpq:0
> Failed to configure the smtp connection:
> [Errno -2] Name or service not known
You have sendemail.smtpserver set to a path, which is supported by
git-sendemail, but not by b4 (yet). I'll see if I can put that in today.
Thank you for testing out b4 send!
-K
Hi,
On Tue, Aug 30, 2022 at 08:22:17AM -0400, Konstantin Ryabitsev wrote:
> (Dropping most cc's, who probably don't care about this thread.)
>
> On Tue, Aug 30, 2022 at 10:30:27AM +0200, Maxime Ripard wrote:
> > > I noticed this, too, and I think the reason is because these patches were
> > > generated with "b4 send -o", but actually sent using git-send-email. It's not
> > > a use-case I've considered (or tested) and the breakage is because when b4
> > > generates patches with "-o", they are written with CRLF line endings, which is
> > > not something git-send-email expects.
> >
> > You're right
>
> OK, the latest master should allow you to do that without causing CRLF errors
> like this series.
Thanks
> > For some reason, b4 send doesn't work for my setup and I haven't had the
> > time to look into why and let you know.
>
> Yeah, I see what the cause of it is:
>
> > Connecting to /home/max/.local/bin/msmtpq:0
> > Failed to configure the smtp connection:
> > [Errno -2] Name or service not known
>
> You have sendemail.smtpserver set to a path, which is supported by
> git-sendemail, but not by b4 (yet). I'll see if I can put that in today.
I saw that you fixed it in master, I'll give it a try, thanks!
Another thing I noticed when using b4 send is that I'm not sure how to
increment the version number of a series. With b4 send --dry-run it
wasn't happening automatically, and I couldn't find anything related in
the options (except for --force-revision but that wasn't permanent I
think?)
I ended up amending the commit json by hand to set the proper revision
number, but I'm almost sure it isn't the right way :)
Maxime
On Wed, Aug 31, 2022 at 09:30:26AM +0200, Maxime Ripard wrote:
> Another thing I noticed when using b4 send is that I'm not sure how to
> increment the version number of a series. With b4 send --dry-run it
> wasn't happening automatically, and I couldn't find anything related in
> the options (except for --force-revision but that wasn't permanent I
> think?)
It is, but I don't blame you being confused seeing as there's no documentation
yet at this point. :) I hope to make some final changes to flags and features
in the near future, so documentation should start appearing soon.
> I ended up amending the commit json by hand to set the proper revision
> number, but I'm almost sure it isn't the right way :)
This is certainly a valid way to go about it as well.
Regards,
-K