Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > linux.kernel > #1511112 > unrolled thread
| Started by | Chris Wilson <chris@chris-wilson.co.uk> |
|---|---|
| First post | 2016-10-28 14:50 +0200 |
| Last post | 2016-10-28 17:00 +0200 |
| Articles | 4 — 3 participants |
Back to article view | Back to linux.kernel
[PATCH] scripts/checkpatch: Check for Reviewed-by under --strict Chris Wilson <chris@chris-wilson.co.uk> - 2016-10-28 14:50 +0200
Re: [Intel-gfx] [PATCH] scripts/checkpatch: Check for Reviewed-by under --strict Jani Nikula <jani.nikula@linux.intel.com> - 2016-10-28 15:40 +0200
Re: [Intel-gfx] [PATCH] scripts/checkpatch: Check for Reviewed-by under --strict Chris Wilson <chris@chris-wilson.co.uk> - 2016-10-28 16:10 +0200
Re: [PATCH] scripts/checkpatch: Check for Reviewed-by under --strict Joe Perches <joe@perches.com> - 2016-10-28 17:00 +0200
| From | Chris Wilson <chris@chris-wilson.co.uk> |
|---|---|
| Date | 2016-10-28 14:50 +0200 |
| Subject | [PATCH] scripts/checkpatch: Check for Reviewed-by under --strict |
| Message-ID | <sxeLE-8o2-5@gated-at.bofh.it> |
Some subsystem polices have a strict requirement that every patch must
have at least one reviewer before being approved for upstream. Since
encouraging review is good policy (great review is even better policy!)
enforce checking for a Reviewed-by when checkpath is run with --strict
(or with --review).
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Andy Whitcroft <apw@canonical.com>
Cc: Joe Perches <joe@perches.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
---
scripts/checkpatch.pl | 17 ++++++++++++++++-
1 file changed, 16 insertions(+), 1 deletion(-)
diff --git a/scripts/checkpatch.pl b/scripts/checkpatch.pl
index a8368d1c4348..9eaa5a4fbbc0 100755
--- a/scripts/checkpatch.pl
+++ b/scripts/checkpatch.pl
@@ -21,6 +21,7 @@ use Getopt::Long qw(:config no_auto_abbrev);
my $quiet = 0;
my $tree = 1;
my $chk_signoff = 1;
+my $chk_review = 0;
my $chk_patch = 1;
my $tst_only;
my $emacs = 0;
@@ -69,6 +70,7 @@ Options:
-q, --quiet quiet
--no-tree run without a kernel tree
--no-signoff do not check for 'Signed-off-by' line
+ --review check for 'Reviewed-by' line
--patch treat FILE as patchfile (default)
--emacs emacs compile window format
--terse one line per report
@@ -183,6 +185,7 @@ GetOptions(
'q|quiet+' => \$quiet,
'tree!' => \$tree,
'signoff!' => \$chk_signoff,
+ 'review!' => \$chk_review,
'patch!' => \$chk_patch,
'emacs!' => \$emacs,
'terse!' => \$terse,
@@ -217,7 +220,7 @@ help(0) if ($help);
list_types(0) if ($list_types);
-$fix = 1 if ($fix_inplace);
+$chk_review = 1 if ($check); # --strict implies checking for Reviewed-by
$check_orig = $check;
my $exit = 0;
@@ -857,6 +860,7 @@ sub git_commit_info {
}
$chk_signoff = 0 if ($file);
+$chk_review = 0 if ($file);
my @rawlines = ();
my @lines = ();
@@ -2130,6 +2134,7 @@ sub process {
our $clean = 1;
my $signoff = 0;
+ my $review = 0;
my $is_patch = 0;
my $in_header_lines = $file ? 0 : 1;
my $in_commit_log = 0; #Scanning lines before patch
@@ -2400,6 +2405,12 @@ sub process {
$in_commit_log = 0;
}
+# Check the patch for any review:
+ if ($line =~ /^\s*reviewed-by:/i) {
+ $review++;
+ $in_commit_log = 0;
+ }
+
# Check if MAINTAINERS is being updated. If so, there's probably no need to
# emit the "does MAINTAINERS need updating?" message on file add/move/delete
if ($line =~ /^\s*MAINTAINERS\s*\|/) {
@@ -6204,6 +6215,10 @@ sub process {
ERROR("MISSING_SIGN_OFF",
"Missing Signed-off-by: line(s)\n");
}
+ if ($is_patch && $has_commit_log && $chk_review && $review == 0) {
+ ERROR("MISSING_REVIEW",
+ "Missing Reviewed-by: line(s)\n");
+ }
print report_dump();
if ($summary && !($clean == 1 && $quiet == 1)) {
--
2.10.1
[toc] | [next] | [standalone]
| From | Jani Nikula <jani.nikula@linux.intel.com> |
|---|---|
| Date | 2016-10-28 15:40 +0200 |
| Subject | Re: [Intel-gfx] [PATCH] scripts/checkpatch: Check for Reviewed-by under --strict |
| Message-ID | <sxfy1-rK-19@gated-at.bofh.it> |
| In reply to | #1511112 |
On Fri, 28 Oct 2016, Chris Wilson <chris@chris-wilson.co.uk> wrote:
> Some subsystem polices have a strict requirement that every patch must
> have at least one reviewer before being approved for upstream. Since
> encouraging review is good policy (great review is even better policy!)
> enforce checking for a Reviewed-by when checkpath is run with --strict
> (or with --review).
Hmm, do you imply the maintainer would have to add his Reviewed-by in
addition to Signed-off-by? I find that a bit too much (especially if you
intend to enforce this over at our corner of the kernel ;)
BR,
Jani.
>
> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> Cc: Andy Whitcroft <apw@canonical.com>
> Cc: Joe Perches <joe@perches.com>
> Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
> ---
> scripts/checkpatch.pl | 17 ++++++++++++++++-
> 1 file changed, 16 insertions(+), 1 deletion(-)
>
> diff --git a/scripts/checkpatch.pl b/scripts/checkpatch.pl
> index a8368d1c4348..9eaa5a4fbbc0 100755
> --- a/scripts/checkpatch.pl
> +++ b/scripts/checkpatch.pl
> @@ -21,6 +21,7 @@ use Getopt::Long qw(:config no_auto_abbrev);
> my $quiet = 0;
> my $tree = 1;
> my $chk_signoff = 1;
> +my $chk_review = 0;
> my $chk_patch = 1;
> my $tst_only;
> my $emacs = 0;
> @@ -69,6 +70,7 @@ Options:
> -q, --quiet quiet
> --no-tree run without a kernel tree
> --no-signoff do not check for 'Signed-off-by' line
> + --review check for 'Reviewed-by' line
> --patch treat FILE as patchfile (default)
> --emacs emacs compile window format
> --terse one line per report
> @@ -183,6 +185,7 @@ GetOptions(
> 'q|quiet+' => \$quiet,
> 'tree!' => \$tree,
> 'signoff!' => \$chk_signoff,
> + 'review!' => \$chk_review,
> 'patch!' => \$chk_patch,
> 'emacs!' => \$emacs,
> 'terse!' => \$terse,
> @@ -217,7 +220,7 @@ help(0) if ($help);
>
> list_types(0) if ($list_types);
>
> -$fix = 1 if ($fix_inplace);
> +$chk_review = 1 if ($check); # --strict implies checking for Reviewed-by
> $check_orig = $check;
>
> my $exit = 0;
> @@ -857,6 +860,7 @@ sub git_commit_info {
> }
>
> $chk_signoff = 0 if ($file);
> +$chk_review = 0 if ($file);
>
> my @rawlines = ();
> my @lines = ();
> @@ -2130,6 +2134,7 @@ sub process {
>
> our $clean = 1;
> my $signoff = 0;
> + my $review = 0;
> my $is_patch = 0;
> my $in_header_lines = $file ? 0 : 1;
> my $in_commit_log = 0; #Scanning lines before patch
> @@ -2400,6 +2405,12 @@ sub process {
> $in_commit_log = 0;
> }
>
> +# Check the patch for any review:
> + if ($line =~ /^\s*reviewed-by:/i) {
> + $review++;
> + $in_commit_log = 0;
> + }
> +
> # Check if MAINTAINERS is being updated. If so, there's probably no need to
> # emit the "does MAINTAINERS need updating?" message on file add/move/delete
> if ($line =~ /^\s*MAINTAINERS\s*\|/) {
> @@ -6204,6 +6215,10 @@ sub process {
> ERROR("MISSING_SIGN_OFF",
> "Missing Signed-off-by: line(s)\n");
> }
> + if ($is_patch && $has_commit_log && $chk_review && $review == 0) {
> + ERROR("MISSING_REVIEW",
> + "Missing Reviewed-by: line(s)\n");
> + }
>
> print report_dump();
> if ($summary && !($clean == 1 && $quiet == 1)) {
--
Jani Nikula, Intel Open Source Technology Center
[toc] | [prev] | [next] | [standalone]
| From | Chris Wilson <chris@chris-wilson.co.uk> |
|---|---|
| Date | 2016-10-28 16:10 +0200 |
| Subject | Re: [Intel-gfx] [PATCH] scripts/checkpatch: Check for Reviewed-by under --strict |
| Message-ID | <sxg13-Tz-7@gated-at.bofh.it> |
| In reply to | #1511136 |
On Fri, Oct 28, 2016 at 04:33:10PM +0300, Jani Nikula wrote: > On Fri, 28 Oct 2016, Chris Wilson <chris@chris-wilson.co.uk> wrote: > > Some subsystem polices have a strict requirement that every patch must > > have at least one reviewer before being approved for upstream. Since > > encouraging review is good policy (great review is even better policy!) > > enforce checking for a Reviewed-by when checkpath is run with --strict > > (or with --review). > > Hmm, do you imply the maintainer would have to add his Reviewed-by in > addition to Signed-off-by? I find that a bit too much (especially if you > intend to enforce this over at our corner of the kernel ;) I do believe we should be keeping the (our, my?) notion of review out of the signed-off-by tag (which imo is a legal statement about the provenance of a patch), and so yes we shouldn't be pushing patches that haven't gone through the rite of fire and been seconded by someone else. -Chris -- Chris Wilson, Intel Open Source Technology Centre
[toc] | [prev] | [next] | [standalone]
| From | Joe Perches <joe@perches.com> |
|---|---|
| Date | 2016-10-28 17:00 +0200 |
| Message-ID | <sxgNx-1cq-7@gated-at.bofh.it> |
| In reply to | #1511112 |
On Fri, 2016-10-28 at 13:49 +0100, Chris Wilson wrote:
> Some subsystem polices have a strict requirement that every patch must
> have at least one reviewer before being approved for upstream. Since
> encouraging review is good policy (great review is even better policy!)
> enforce checking for a Reviewed-by when checkpath is run with --strict
> (or with --review).
I rather dislike this as it imposes a rule outside
of what's documented in SubmittingPatches.
Ideally, please keep a private version of this.
And unless and until SubmittingPatches is updated,
please keep this separate from --strict.
> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> Cc: Andy Whitcroft <apw@canonical.com>
> Cc: Joe Perches <joe@perches.com>
> Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
> ---
> scripts/checkpatch.pl | 17 ++++++++++++++++-
> 1 file changed, 16 insertions(+), 1 deletion(-)
>
> diff --git a/scripts/checkpatch.pl b/scripts/checkpatch.pl
> index a8368d1c4348..9eaa5a4fbbc0 100755
> --- a/scripts/checkpatch.pl
> +++ b/scripts/checkpatch.pl
> @@ -21,6 +21,7 @@ use Getopt::Long qw(:config no_auto_abbrev);
> my $quiet = 0;
> my $tree = 1;
> my $chk_signoff = 1;
> +my $chk_review = 0;
> my $chk_patch = 1;
> my $tst_only;
> my $emacs = 0;
> @@ -69,6 +70,7 @@ Options:
> -q, --quiet quiet
> --no-tree run without a kernel tree
> --no-signoff do not check for 'Signed-off-by' line
> + --review check for 'Reviewed-by' line
> --patch treat FILE as patchfile (default)
> --emacs emacs compile window format
> --terse one line per report
> @@ -183,6 +185,7 @@ GetOptions(
> 'q|quiet+' => \$quiet,
> 'tree!' => \$tree,
> 'signoff!' => \$chk_signoff,
> + 'review!' => \$chk_review,
> 'patch!' => \$chk_patch,
> 'emacs!' => \$emacs,
> 'terse!' => \$terse,
> @@ -217,7 +220,7 @@ help(0) if ($help);
>
> list_types(0) if ($list_types);
>
> -$fix = 1 if ($fix_inplace);
> +$chk_review = 1 if ($check); # --strict implies checking for Reviewed-by
> $check_orig = $check;
>
> my $exit = 0;
> @@ -857,6 +860,7 @@ sub git_commit_info {
> }
>
> $chk_signoff = 0 if ($file);
> +$chk_review = 0 if ($file);
>
> my @rawlines = ();
> my @lines = ();
> @@ -2130,6 +2134,7 @@ sub process {
>
> our $clean = 1;
> my $signoff = 0;
> + my $review = 0;
> my $is_patch = 0;
> my $in_header_lines = $file ? 0 : 1;
> my $in_commit_log = 0; #Scanning lines before patch
> @@ -2400,6 +2405,12 @@ sub process {
> $in_commit_log = 0;
> }
>
> +# Check the patch for any review:
> + if ($line =~ /^\s*reviewed-by:/i) {
> + $review++;
> + $in_commit_log = 0;
> + }
> +
> # Check if MAINTAINERS is being updated. If so, there's probably no need to
> # emit the "does MAINTAINERS need updating?" message on file add/move/delete
> if ($line =~ /^\s*MAINTAINERS\s*\|/) {
> @@ -6204,6 +6215,10 @@ sub process {
> ERROR("MISSING_SIGN_OFF",
> "Missing Signed-off-by: line(s)\n");
> }
> + if ($is_patch && $has_commit_log && $chk_review && $review == 0) {
> + ERROR("MISSING_REVIEW",
> + "Missing Reviewed-by: line(s)\n");
> + }
>
> print report_dump();
> if ($summary && !($clean == 1 && $quiet == 1)) {
[toc] | [prev] | [standalone]
Back to top | Article view | linux.kernel
csiph-web