On 9/4/25 10:30 AM, Philip Oberfichtner wrote:
On Wed, Sep 03, 2025 at 03:02:20PM +0200, Marek Vasut wrote:
On 9/3/25 12:56 PM, Philip Oberfichtner wrote:
Deduplicate some redundancies within the SIZE_CHECK code in the toplevel
Makefile. No functional changes.
Note: To make this work for the special case of SPL, spl_size_limit.c
has to print an empty string "" instead of zero when there is no size
limit in place.
Signed-off-by: Philip Oberfichtner <p...@denx.de>
---
Makefile | 45 +++++++++++-------------------------------
tools/spl_size_limit.c | 4 ++--
2 files changed, 13 insertions(+), 36 deletions(-)
diff --git a/Makefile b/Makefile
index a17d8b57196..b6f01591ee9 100644
--- a/Makefile
+++ b/Makefile
@@ -428,8 +428,9 @@ KBUILD_HOSTLDLIBS := $(HOST_LFS_LIBS) $(HOSTLDLIBS)
# Check ths size of a binary:
# Args:
# $1: File to check
-# #2: Size limit in bytes (decimal or 0xhex)
+# #2: Size limit in bytes (decimal or 0xhex), always return OK if undefined
You should be able to use the HAS_ variables to correctly test for and
handle the undefined case .
In general, I agree. But this would mean to reintroduce the multiply
duplicated Makefile bloat à la:
ifeq ($(CONFIG_HAS_BOARD_SIZE_LIMIT),y)
BOARD_SIZE_CHECK= @ $(call size_check,$@,$(CONFIG_BOARD_SIZE_LIMIT))
else
BOARD_SIZE_CHECK =
endif
That's why after all I preferred my "test -z" solution. Is there maybe
a third way I am missing out on?
What do you think about giving size_check another argument, i.e.
call size_check,$@,$(CONFIG_HAS_BOARD_SIZE_LIMIT),$(CONFIG_BOARD_SIZE_LIMIT)
This could be an option , yes .
