linux-next: Signed-off-by missing for commits in Linus' tree

linux-next: Signed-off-by missing for commits in Linus' tree

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 709 bytes --]

Hi all,

Commits

  b33fff07e3e3 ("x86, build: allow LTO to be selected")
  d2dcd3e37475 ("x86, cpu: disable LTO for cpu.c")
  e242db40be27 ("x86, vdso: disable LTO only for vDSO")
  b1a1a1a09b46 ("kbuild: lto: postpone objtool")
  41425ebe2024 ("objtool: Split noinstr validation from --vmlinux")
  6dafca978033 ("x86, build: use objtool mcount")
  22c8542d7b22 ("tracing: add support for objtool mcount")
  0e731dbc1824 ("objtool: Don't autodetect vmlinux.o")
  18a14575ae31 ("objtool: Fix __mcount_loc generation with Clang's assembler")
  99d0021569c7 ("objtool: Add a pass for generating __mcount_loc")

are missing a Signed-off-by from their committer.

-- 
Cheers,
Stephen Rothwell

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

Re: linux-next: Signed-off-by missing for commits in Linus' tree

[Linus Torvalds]

On Tue, Feb 23, 2021 at 4:31 PM Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>
> are missing a Signed-off-by from their committer.

Gaah. Maybe I should do some pre-pull hook or something to notice this
automatically (like you clearly do).

            Linus

Re: linux-next: Signed-off-by missing for commits in Linus' tree

[Stephen Rothwell]

[-- Attachment #1.1: Type: text/plain, Size: 585 bytes --]

Hi Linus,

On Tue, 23 Feb 2021 16:33:47 -0800 Linus Torvalds <torvalds@linux-foundation.org> wrote:
>
> On Tue, Feb 23, 2021 at 4:31 PM Stephen Rothwell <sfr@canb.auug.org.au> wrote:
> >
> > are missing a Signed-off-by from their committer.  
> 
> Gaah. Maybe I should do some pre-pull hook or something to notice this
> automatically (like you clearly do).

I have attached the scripts I run over things after fetching them, but
before merging them (so not a hook, sorry).  check_commits runs
check_fixes - but just for my convenience.

-- 
Cheers,
Stephen Rothwell

[-- Attachment #1.2: check_commits --]
[-- Type: application/x-shellscript, Size: 1611 bytes --]

[-- Attachment #1.3: check_fixes --]
[-- Type: application/x-shellscript, Size: 4299 bytes --]

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

Re: linux-next: Signed-off-by missing for commits in Linus' tree

[Kees Cook]

On Wed, Feb 24, 2021 at 11:49:42AM +1100, Stephen Rothwell wrote:
> Hi Linus,
> 
> On Tue, 23 Feb 2021 16:33:47 -0800 Linus Torvalds <torvalds@linux-foundation.org> wrote:
> >
> > On Tue, Feb 23, 2021 at 4:31 PM Stephen Rothwell <sfr@canb.auug.org.au> wrote:
> > >
> > > are missing a Signed-off-by from their committer.  
> > 
> > Gaah. Maybe I should do some pre-pull hook or something to notice this
> > automatically (like you clearly do).
> 
> I have attached the scripts I run over things after fetching them, but
> before merging them (so not a hook, sorry).  check_commits runs
> check_fixes - but just for my convenience.

Thank you! I've added these to my PR workflow now, and it yells quite loudly.
I'm still looking at some kind of push hook too...

-- 
Kees Cook

Re: linux-next: Signed-off-by missing for commits in Linus' tree

[Kees Cook]

On Wed, Feb 24, 2021 at 11:49:42AM +1100, Stephen Rothwell wrote:
> Hi Linus,
> 
> On Tue, 23 Feb 2021 16:33:47 -0800 Linus Torvalds <torvalds@linux-foundation.org> wrote:
> >
> > On Tue, Feb 23, 2021 at 4:31 PM Stephen Rothwell <sfr@canb.auug.org.au> wrote:
> > >
> > > are missing a Signed-off-by from their committer.  
> > 
> > Gaah. Maybe I should do some pre-pull hook or something to notice this
> > automatically (like you clearly do).
> 
> I have attached the scripts I run over things after fetching them, but
> before merging them (so not a hook, sorry).  check_commits runs
> check_fixes - but just for my convenience.
> 
> -- 
> Cheers,
> Stephen Rothwell

Based on the pre-push.sample file, here is what I've added to my
.git/hooks/pre-push hook:


z40=0000000000000000000000000000000000000000

while read local_ref local_sha remote_ref remote_sha
do
        if [ "$local_sha" = $z40 ]
        then
                # Handle delete
                :
        else
                if [ "$remote_sha" = $z40 ]
                then
                        # New branch, examine all commits
                        range="$local_sha"
                else
                        # Update to existing branch, examine new commits
                        range="$remote_sha..$local_sha"
                fi

                if ! $HOME/bin/check_commits "$range"; then
                        exit 1
                fi
        fi
done

(and I modified check_fixes and check_commits to exit non-zero on failure)


I wonder if we need this in Documentation/maintainer/configure-git.rst
and to put check_commits and check_fixes into tools/ somewhere?

(Though goodness, please never aim your hook at your tree's tools/
directory.)

-- 
Kees Cook

Re: linux-next: Signed-off-by missing for commits in Linus' tree

[Linus Walleij]

On Wed, Feb 24, 2021 at 10:30 PM Kees Cook <keescook@chromium.org> wrote:

> I wonder if we need this in Documentation/maintainer/configure-git.rst

That is generally a good idea.

> and to put check_commits and check_fixes into tools/ somewhere?

People keep saying we should put git hooks somewhere where
they can be reused easily. What about just creating
scripts/git-hooks?

Yours,
Linus Walleij

Re: linux-next: Signed-off-by missing for commits in Linus' tree

[Kees Cook]

On Wed, Feb 24, 2021 at 11:31:08AM +1100, Stephen Rothwell wrote:
> Hi all,
> 
> Commits
> 
>   b33fff07e3e3 ("x86, build: allow LTO to be selected")
>   d2dcd3e37475 ("x86, cpu: disable LTO for cpu.c")
>   e242db40be27 ("x86, vdso: disable LTO only for vDSO")
>   b1a1a1a09b46 ("kbuild: lto: postpone objtool")
>   41425ebe2024 ("objtool: Split noinstr validation from --vmlinux")
>   6dafca978033 ("x86, build: use objtool mcount")
>   22c8542d7b22 ("tracing: add support for objtool mcount")
>   0e731dbc1824 ("objtool: Don't autodetect vmlinux.o")
>   18a14575ae31 ("objtool: Fix __mcount_loc generation with Clang's assembler")
>   99d0021569c7 ("objtool: Add a pass for generating __mcount_loc")
> 
> are missing a Signed-off-by from their committer.

Ieeeee! Ugh, yes, my bad, entirely. I screwed up when rebuilding the
LTO "part 2" series for the -rc1 window (missed the -s on the cherry-pick).

Since we can't change git history, the best fix I can do is send it here
to the list. Obviously, these should all be considered:

Signed-off-by: Kees Cook <keescook@chromium.org>

Again, apologies. I, too, will add a commit hook to check for this. (Is
there an existing script anyone else is already using?)

-- 
Kees Cook