[PATCH v5 1/8] binman: add support for skipping file concatenation for mkimage

[Simon Glass]

Hi Quentin,

On Tue, 30 Aug 2022 at 03:57, Quentin Schulz
<quentin.schulz at theobroma-systems.com> wrote:
>
> Hi Simon,
>
> On 8/27/22 02:21, Simon Glass wrote:
> > Hi Quentin,
> >
> > On Fri, 26 Aug 2022 at 09:37, Quentin Schulz <foss+uboot at 0leil.net> wrote:
> >>
> >> From: Quentin Schulz <quentin.schulz at theobroma-systems.com>
> >>
> >> Some image types handled by mkimage require the datafiles to be passed
> >> independently (-d data1:data2) for specific handling of each. A
> >> concatenation of datafiles prior to passing them to mkimage wouldn't
> >> work.
> >>
> >> That is the case for rkspi for example which requires page alignment
> >> and only writing 2KB every 4KB.
> >>
> >> This adds the ability to tell binman to pass the datafiles without
> >> prior concatenation to mkimage, by adding the multiple-data-files
> >> boolean property to the mkimage node.
> >>
> >> Cc: Quentin Schulz <foss+uboot at 0leil.net>
> >> Reviewed-by: Simon Glass <sjg at chromium.org>
> >> Signed-off-by: Quentin Schulz <quentin.schulz at theobroma-systems.com>
> >> ---
> >>
> >> v5:
> >>   - changed to use full path from input dir with tools.get_input_filename
> >>   to make it possible to run the unit tests,
> >>   - added unit test,
> >>
> >>
> >>   tools/binman/entries.rst                      | 22 ++++++++++
> >>   tools/binman/etype/mkimage.py                 | 41 +++++++++++++++++--
> >>   tools/binman/ftest.py                         | 16 ++++++++
> >
> > Please put the new test at the end.
> >
> >>   .../test/241_mkimage_multiple_data_files.dts  | 21 ++++++++++
> >>   4 files changed, 96 insertions(+), 4 deletions(-)
> >>   create mode 100644 tools/binman/test/241_mkimage_multiple_data_files.dts
> >
> > This is pretty close but it still missing a line of test coverage.
> > Please try 'binman test -T' to see it. I'd also prefer a shorter
>
> This does not work on Fedora.
> 1) there's no python3-coverage binary available,
> 2) After replacing python3-coverage with just coverage, the tests are
> stuck and never finish, (I have seen the patches to use COVERAGE
> environment variable so I guess the required changes might be tackled
> soon in master),
>
> Any tip on how to identify which test is stuck except going through them
> one by one?

One way is to add comment blocks '''...''' across the ftest.py file,
using a binary chop to identify the problem.

Or, since tests are run in series, you could hack test_util to pass
verbose parameters when it runs the tests - see 'cmd =' in
run_test_coverage().

>
> python3-coverage is also not available in the container image built from
> tools/docker/Dockerfile.

does 'python3 -m coverage' work?

or this:

https://coverage.readthedocs.io/en/6.3.2/install.html

>
> > filename for the 241 file.
> >
> > I've pushed a tree containing a suggested fix (updating this patch). I
> > can update it when applying if you like, otherwise please send a new
> > version.
> >
>
> Where did you push the tree?

Sorry I forgot to mention that:

https://github.com/sjg20/u-boot/tree/try-rk4

Regards,
SImon