Re: [PATCH resend] media: staging: stkwebcam: Restore MEDIA_{USB,CAMERA}_SUPPORT dependencies

[Geert Uytterhoeven]

Hi Dan,

On Wed, Nov 23, 2022 at 4:52 PM Dan Carpenter <error27@xxxxxxxxx> wrote:
> On Wed, Nov 23, 2022 at 04:49:05PM +0100, Geert Uytterhoeven wrote:
> > On Wed, Nov 23, 2022 at 4:11 PM Dan Carpenter <error27@xxxxxxxxx> wrote:
> > > On Wed, Nov 23, 2022 at 11:13:31AM +0100, Geert Uytterhoeven wrote:
> > > > On Wed, Nov 23, 2022 at 11:08 AM Tommaso Merciai
> > > > <tommaso.merciai@xxxxxxxxxxxxxxxxxxxx> wrote:
> > > > > On Mon, Nov 21, 2022 at 04:58:33PM +0100, Geert Uytterhoeven wrote:
> > > > > > By moving support for the USB Syntek DC1125 Camera to staging, the
> > > > > > dependencies on MEDIA_USB_SUPPORT and MEDIA_CAMERA_SUPPORT were lost.
> > > > > >
> > > > > > Fixes: 56280c64ecacc971 ("media: stkwebcam: deprecate driver, move to staging")
> > > > >
> > > > > Patch itself looks good but we have some style issue. Applying this
> > > > > patch I got the following warning from checkpatchl:
> > > > >
> > > > > WARNING: Please use correct Fixes: style 'Fixes: <12 chars of sha1> ("<title line>")' - ie: 'Fixes: 56280c64ecac ("media: stkwebcam: deprecate driver, move to staging")'
> > > > > #10:
> > > > >
> > > > > You have to pass only the first 12 chars of the sha1 commit into Fixes
> > > > > msg:
> > > > >
> > > > > Use:
> > > > >
> > > > >  Fixes: 56280c64ecac ("media: stkwebcam: deprecate driver, move to staging")
> > > > >
> > > > > Instead of:
> > > > >
> > > > >  Fixes: 56280c64ecacc971 ("media: stkwebcam: deprecate driver, move to staging")
> > > >
> > > > I always use 16 chars, to avoid these becoming ambiguous in a few years.
> > >
> > > If we assume hashes are randomly distributed and that people commit
> > > 100k patches every year then with 12 character we would have 17
> > > collisions every 1000 years.
> >
> > So I can expect to see a collision before my retirement day
> > (which coincides with the signed 32-bit time_t flag day ;-)
> >
> > BTW, does the above take into account that commit hashes can
> > collide with other object type hashes, too?
>
> I assumed that `git show` won't show those other object types, but I
> don't really know if that's true.

"git show" is not limited to commit types:

$ git cat-file commit eb7081409f94a9a8608593d0fb63a1aa3d6f95d8
tree 0e320b5ae477efe1c9928057762bf63d730204ce
parent c6c67bf9bc2714d9c2c2e7ecfbf29d912b8c4f17
author Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx> 1668988936 -0800
committer Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx> 1668988936 -0800

Linux 6.1-rc6
$ git show 0e320b5ae477efe1c9928057762bf63d730204ce
tree 0e320b5ae477efe1c9928057762bf63d730204ce

.clang-format
.cocciconfig
[...]
$ git show eb708
error: short SHA1 eb708 is ambiguous
hint: The candidates are:
hint:   eb7081409f94a9a8 commit 2022-11-20 - Linux 6.1-rc6
hint:   eb708b0ff972bfe0 commit 2016-11-12 - arm64: dts: Add ARM PMU
node for exynos7
hint:   eb7080213d0fee54 tree
hint:   eb7080d36f660012 tree
hint:   eb7083e15bcb5eea tree
hint:   eb7083f45a088bf3 tree
hint:   eb7084a7ae5f486a tree
hint:   eb70866f5c60a477 tree
hint:   eb7087c754052a42 tree
hint:   eb7089eeb5f67a5f tree
hint:   eb708a906ceeea91 tree
hint:   eb708c1853ce1355 tree
hint:   eb708ca4d77be915 tree
hint:   eb708e4839466fdd tree
hint:   eb708431cb964d36 blob
hint:   eb708b77c4a54a68 blob
fatal: ambiguous argument 'eb708': unknown revision or path not in the
working tree.
Use '--' to separate paths from revisions, like this:
'git <command> [<revision>...] -- [<file>...]'

I couldn't find an easy way to count the number of objects per
type in a repository.

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@xxxxxxxxxxxxxx

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds