Hi Tom,
On Thu, 22 Nov 2018 at 16:16, Tom Rini trini@konsulko.com wrote:
On Thu, Nov 22, 2018 at 01:09:16PM -0700, Simon Glass wrote:
The deadline for this has passed and we are starting to remove boards that do not use driver model for block devices.
Add a noisy Makefile warning.
Signed-off-by: Simon Glass sjg@chromium.org
Makefile | 7 +++++++ 1 file changed, 7 insertions(+)
diff --git a/Makefile b/Makefile index 552687db538..8bf422f6a85 100644 --- a/Makefile +++ b/Makefile @@ -917,6 +917,13 @@ ifeq ($(CONFIG_DM_I2C_COMPAT)$(CONFIG_SANDBOX),y) @echo "(possibly in a subsequent patch in your series)" @echo "before sending patches to the mailing list." @echo "====================================================" +endif +ifneq ($(CONFIG_BLK),y)
@echo "===================== WARNING ======================"@echo "This board does not use CONFIG_BLK. Please update"@echo "the board to use CONFIG_BLK before the end of 2018."@echo "See doc/driver-model/MIGRATION.txt for more info."@echo "===================================================="endif
The problem with this is that it doesn't catch most of the problem combinations. I'm still working on a series that highlights what the problem cases we have left are. But it's also not, in a lot of cases, a board problem, but a lack of driver being updated problem so I suspect we'll really be marking a bunch of stuff as depends on BROKEN.
Are you saying that CONFIG_BLK is set to y but the board is still not converted? I'm not sure I understand the issue here.
Of course, one problem here is that CONFIG_BLK might not be needed if the board doesn't use block devices.
Regards, Simon