warning: no case matching constant switch condition '0'

[nickdesaulniers]

drivers/block/drbd/* has tons of warnings like:

In file included from drivers/block/drbd/drbd_actlog.c:30:
In file included from drivers/block/drbd/drbd_int.h:48:
In file included from ./include/linux/drbd_genl_api.h:54:
In file included from ./include/linux/genl_magic_struct.h:236:
./include/linux/drbd_genl.h:317:1: warning: no case matching constant switch condition
      '0'
GENL_struct(DRBD_NLA_NOTIFICATION_HEADER, 23, drbd_notification_header,
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
./include/linux/genl_magic_struct.h:220:10: note: expanded from macro 'GENL_struct'
        switch (0) {                                                    \
                ^

Without looking at the source code for more context switch(0) already looks hyper-wrong (if (false)), but maybe there's another way to accomplish whatever they're using that construct for. Probably needs CONFIG_BLK_DEV_DRBD=y.

[nickdesaulniers]

Seems to have been added in commit ec2c35a ("drbd: prepare the transition from connector to genetlink")

[bulwahn]

Philipp Klocke resend Arnd Bergmann's patch that was not merged inn 2017 to resolve this issue here:
https://lore.kernel.org/r/4ffc2255-2ae2-60d4-703e-151979efe7ab@gmx.de/

[nathanchance]

That patch should have been CC'd to Jens and linux-block since he's been taking drbd patches for a while now. I'll resend that patch (since the attribution is wrong on it) and my patch from #54 tomorrow unless there are any objections.

I also think that file should be added to the drbd section in MAINTAINERS so it's clear who should be taking the patches (since drbd introduced it and has been the only consumer since then), I'll ask to see what they think about that when I resent.

[bulwahn]

Sure, go ahead and do so. My student Philipp is currently on vacation, and the patch is authored by Arnd Bergmann anyway.

[nathanchance]

Patch resent: https://lore.kernel.org/lkml/20181002171447.10765-1-natechancellor@gmail.com/

[nickdesaulniers]

Thanks for the resend @nathanchance .

@bulwahn sounds like you're a teacher? Anything we can do to help involve students in kernel or llvm development?

[bulwahn]

@nickdesaulniers I am actually just a software engineer, probably at the moment more in the role of a software project manager. I am interested in Linux kernel development, and try to offer students the possibility to work on the Linux kernel, paying them to read code, find issues, use code analysis and write patches.

I think your blog motivated to get started around some activities with the kernel and clang. This github issue tracker fits very well to keep track of who is currently involved with which issue and collects the state of discussion quite well. Thanks for this effort!

Other than keeping up this good work, I do currently not see much other things to be done. I might eventually report on my experience with running Facebook's Infer static analyser on the kernel, and another supporting person is planned to look at the clang-tidy results in bit more in detail, which I can also report on once we have some actual deeper insights.

[nickdesaulniers]

paying them to read code, find issues, use code analysis and write patches.

Out of pocket? ⚡

I think your blog motivated

Well then writing it wasn't a complete wash. 💇‍♂️ I haven't blogged too much about the Clang built kernel stuff, probably should.

I might eventually report on my experience with running Facebook's Infer static analyser on the kernel

This whole effort started because I wanted to run Clang's static analyzer on the kernel, and started picking up other dependencies until building the kernel outright became the predominate blocker to so much.

Other than keeping up this good work

I appreciate everyone's help on this, including yours.

[nathanchance]

Patch v2 with Lars' suggested change: https://lore.kernel.org/lkml/20181004200912.32348-1-natechancellor@gmail.com/

[dileks]

@nathanchance Suggestions for a v3:

In [1]:
Subject: s/pointless switch statment/pointless switch statement (missing "e" in statement)

If you send out a v3 you can add Arnd's "Fixes:" tag, too.
From [2]:
Fixes: ec2c35a ("drbd: prepare the transition from connector to genetlink")

[1] https://lore.kernel.org/r/20181004200912.32348-1-natechancellor@gmail.com/
[2] https://lore.kernel.org/r/20170201165532.2605534-1-arnd@arndb.de/

[nathanchance]

Patch accepted: https://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux-block.git/commit/?h=for-4.21/block-late&id=a52c5a16cf19d8a85831bb1b915a221dd4ffae3c

Will be in the later block pull request for 4.21/5.0 because the drbd pull request happened too late: https://lore.kernel.org/lkml/84bdf9f3-c5f5-6192-703b-c1608f725ba3@kernel.dk/

[nathanchance]

Merged into mainline: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=a52c5a16cf19d8a85831bb1b915a221dd4ffae3c