| commit | 3e51d530645059093e6dd27c4bbfafb8a216cd41 | |
| author | Nico Huber <nico.h@gmx.de> | |
| Fri, 17 Aug 2018 18:05:53 +0000 (17 20:05 +0200) | ||
| committer | Patrick Georgi <pgeorgi@google.com> | |
| Thu, 13 Sep 2018 13:58:29 +0000 (13 13:58 +0000) | ||
| tree | eba72a05a204b276af0b7771dcb2330fd8286ae8 | treesnapshot (tar.gz zip) |
| parent | 392e886552afa5c9d33bc443290c7c72303b6c34 | commitdiff |
Makefile.inc: Fix dependency tracking of fmap{_config.h,.desc}
GNU make is too smart (or too stupid?) for empty recipes. In the case of
empty recipes, GNU make doesn't consider the target as updated even if
its prerequisites are. So if we told make to rebuild `build/romstage/
lib/cbfs.o` for instance, and the FMAP changed, it rerun the fmaptool
recipe (as a prerequisite) but only considered `cbfs.o` to be updated
by chance.
Just not leaving the recipes empty seems to help here. I seeemed to
remember that it wasn't that easy, but it fixes the issue for me...
Change-Id: Ic7ecb88cf7df7f2488defd47ea02255fc10a67e9
Signed-off-by: Nico Huber <nico.h@gmx.de>
Reviewed-on: https://review.coreboot.org/28198
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
GNU make is too smart (or too stupid?) for empty recipes. In the case of
empty recipes, GNU make doesn't consider the target as updated even if
its prerequisites are. So if we told make to rebuild `build/romstage/
lib/cbfs.o` for instance, and the FMAP changed, it rerun the fmaptool
recipe (as a prerequisite) but only considered `cbfs.o` to be updated
by chance.
Just not leaving the recipes empty seems to help here. I seeemed to
remember that it wasn't that easy, but it fixes the issue for me...
Change-Id: Ic7ecb88cf7df7f2488defd47ea02255fc10a67e9
Signed-off-by: Nico Huber <nico.h@gmx.de>
Reviewed-on: https://review.coreboot.org/28198
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
| Makefile.inc | diffblobblamehistory |