Re: [PATCH v2 2/5] livepatch/klp-build: handle patches that add/remove files

[Joe Lawrence <joe.lawrence@redhat.com>]

On Thu, Feb 05, 2026 at 08:53:42AM -0800, Josh Poimboeuf wrote:
> On Thu, Feb 05, 2026 at 11:35:00AM -0500, Joe Lawrence wrote:
> > On Wed, Feb 04, 2026 at 10:02:38AM -0800, Josh Poimboeuf wrote:
> > > On Tue, Feb 03, 2026 at 09:51:37PM -0500, Joe Lawrence wrote:
> > > > The klp-build script prepares a clean patch by populating two temporary
> > > > directories ('a' and 'b') with source files and diffing the result.
> > > > However, this process currently fails when a patch introduces a new
> > > > source file as the script attempts to copy files that do not yet exist
> > > > in the original source tree.  Likewise, there is a similar limitation
> > > > when a patch removes a source file and the script tries to copy files
> > > > that no longer exist.
> > > > 
> > > > Refactor the file-gathering logic to distinguish between original input
> > > > files and patched output files:
> > > > 
> > > > - Split get_patch_files() into get_patch_input_files() and
> > > >   get_patch_output_files() to identify which files exist before and
> > > >   after patch application.
> > > > - Filter out "/dev/null" from both to handle file creation/deletion
> > > > - Update refresh_patch() to only copy existing input files to the 'a'
> > > >   directory and the resulting output files to the 'b' directory.
> > > > 
> > > > This allows klp-build to successfully process patches that add or remove
> > > > source files.
> > > > 
> > > > Signed-off-by: Joe Lawrence <joe.lawrence@redhat.com>
> > > > ---
> > > >  scripts/livepatch/klp-build | 34 +++++++++++++++++++++++++++-------
> > > >  1 file changed, 27 insertions(+), 7 deletions(-)
> > > > 
> > > > Lightly tested with patches that added or removed a source file, as
> > > > generated by `git diff`, `git format-patch`, and `diff -Nupr`.
> > > > 
> > > > diff --git a/scripts/livepatch/klp-build b/scripts/livepatch/klp-build
> > > > index 9f1b77c2b2b7..5a99ff4c4729 100755
> > > > --- a/scripts/livepatch/klp-build
> > > > +++ b/scripts/livepatch/klp-build
> > > > @@ -299,15 +299,33 @@ set_kernelversion() {
> > > >  	sed -i "2i echo $localversion; exit 0" scripts/setlocalversion
> > > >  }
> > > >  
> > > > -get_patch_files() {
> > > > +get_patch_input_files() {
> > > > +	local patch="$1"
> > > > +
> > > > +	grep0 -E '^--- ' "$patch"				\
> > > > +		| gawk '{print $2}'				\
> > > > +		| grep -v '^/dev/null$'				\
> > > 
> > > Because pipefail is enabled, the grep0 helper should be used instead of
> > > grep, otherwise a failed match can propagate to an error.  Maybe we need
> > > a "make check" or something which enforces that and runs shellcheck.
> > > 

How about defining our own grep in the script that intercepts the call
and throws an error:

  grep() {
  	echo "error: $SCRIPT: use grep0 or 'command grep' instead of bare grep" >&2
  	exit 1
  }

That seems easier than trying to externally parse the script to figure
out what's a command, comment, word-match, legit grep, etc.

--
Joe