On Mon, 29 Sept 2025 at 10:16, Bruce Richardson <[email protected]> wrote: > > On Fri, Sep 26, 2025 at 02:41:01PM +0200, David Marchand wrote: > > A problem with the current headers check is that it relies on > > meson dependencies objects that come with their include_directories > > directives, and all of those point at the library / driver sources. > > > > This means that we won't detect a public header including a private > > (as in, not exported) header, or a driver only header. > > > > To address this issue, a staging directory is added and every header > > is copied to it. > > > > Drivers and library headers are staged to two different directories > > and the check is updated accordingly. > > > > Signed-off-by: David Marchand <[email protected]> > > --- > > Changes since v4: > > - split global_inc (isolating config include path from EAL headers), > > - moved all changes under buildtools/chkincs, > > > > Changes since v3: > > - removed update of global_inc (which was unneeded, and probably was > > what triggered a Windows clang build issue reported by CI), > > - moved all staging operations under a check on check_includes= option, > > - moved staging directories under buildtools/chkincs/, > > - renamed all variables to reflect those concern the headers check, > > - made chkincs binaries depend on staging deps instead of having the > > libraries/drivers depend on them, > > - added C++ check for driver headers (missed in v3), > > > > --- > > > Played around a bit last week with options for doing chkincs, however, I > didn't find anything obviously simpler or better than this. Let's go with > this solution! > > Acked-by: Bruce Richardson <[email protected]>
Thanks for the review. I got an interesting comment from a rabbit (in the middle of a lot of noise ;-)). https://github.com/ovsrobot/dpdk/pull/106#discussion_r2382870892 That's not a common case, but this could be tricky to someone moving around/renaming headers. What do you think? -- David Marchand
