[PATCH 00/49] image: Reduce #ifdefs and ad-hoc defines in image code

[Tom Rini]

On Tue, May 04, 2021 at 07:24:25PM -0400, Sean Anderson wrote:
> Hi Tom,
> 
> On 5/4/21 5:40 PM, Tom Rini wrote:
> > On Mon, May 03, 2021 at 05:10:47PM -0600, Simon Glass wrote:
> > 
> > > Much of the image-handling code predates the introduction of Kconfig and
> > > has quite a few #ifdefs in it. It also uses its own IMAGE_... defines to
> > > help reduce the #ifdefs, which is unnecessary now that we can use
> > > IS_ENABLED() et al.
> > > 
> > > The image code is also where quite a bit of code is shared with the host
> > > tools. At present this uses a lot of checks of USE_HOSTCC.
> > > 
> > > This series introduces 'host' Kconfig options and a way to use
> > > CONFIG_IS_ENABLED() to check them. This works in a similar way to SPL, so
> > > 
> > >     CONFIG_IS_ENABLED(FIT)
> > > 
> > > will evaluate to true on the host build (USE_HOSTCC) if CONFIG_HOST_FIT is
> > > enabled. This allows quite a bit of clean-up of the image.h header file
> > > and many of the image C files.
> > > 
> > > The 'host' Kconfig options should help to solve a more general problem in
> > > that we mostly want the host tools to build with all features enabled, no
> > > matter which features the 'target' build actually uses. This is a pain to
> > > arrange at present, but with 'host' Kconfigs, we can just define them all
> > > to y.
> > > 
> > > There are cases where the host tools do not have features which are
> > > present on the target, for example environment and physical addressing.
> > > To help with this, some of the core image code is split out into
> > > image-board.c and image-host.c files.
> > > 
> > > Even with these changes, some #ifdefs remain (101 down to 42 in
> > > common/image*). But the code is somewhat easier to follow and there are
> > > fewer build paths.
> > > 
> > > In service of the above, this series includes a patch to add an API function
> > > for zstd, so the code can be dropped from bootm.c
> > > 
> > > It also introduces a function to handle manual relocation.
> > 
> > I like this idea overall.  The good news is this reduces the size in a
> > few places.  The bad news, but I can live with if we can't restructure
> > the changes more, is a few functions grow a bit.  This shows the good
> > and the bad (something like sama5d2_ptc_ek_mmc shows only growth, to be
> > clear):
> What tool do you use to generate this output? Thanks.

buildman will give that for you.  This was from my world build
before/after wrapper, but for a single machine I have:
#!/bin/bash

# Initial and constant buildman args
ARGS="-devl"
ALL=0
KEEP=0

# Find our arguments
while test $# -ne 0; do
	if [ "$1" == "--all" ]; then
		ALL=1
		shift 1
	elif [ "$1" == "--branch" ]; then
		BRANCH=$2
		shift 2
	elif [ "$1" == "--keep" ]; then
		KEEP=1
		ARGS="$ARGS -k"
		shift 1
	else
		MACHINE=$1
		shift
	fi
done

if [ -z $MACHINE ]; then
	echo Usage: $0 MACHINE [--all] [--keep] [--branch BRANCH]
	exit 1
fi

# If not all, then only first/last
if [ $ALL -ne 1 ]; then
	ARGS="$ARGS --step 0"
fi

if [ ! -z $BRANCH ]; then
	ARGS="$ARGS -b $BRANCH"
else
	ARGS="$ARGS -b `git rev-parse --abbrev-ref HEAD`"
fi

mkdir -p /tmp/$MACHINE

export SOURCE_DATE_EPOCH=`date +%s`
./tools/buildman/buildman -o /tmp/$MACHINE $ARGS -SBC $MACHINE
./tools/buildman/buildman -o /tmp/$MACHINE $ARGS -SsB $MACHINE

[ $KEEP -eq 0 ] && rm -rf /tmp/$MACHINE

In a script called "uboot-size-test.sh" to dig in to individual cases
more.

-- 
Tom
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 659 bytes
Desc: not available
URL: <https://lists.denx.de/pipermail/u-boot/attachments/20210504/69e52d4b/attachment.sig>