[PATCH 1/3] checkpatch.pl: add missing memory barriers

[PATCH 1/3] checkpatch.pl: add missing memory barriers

[Michael S. Tsirkin]

SMP-only barriers were missing in checkpatch.pl

Refactor code slightly to make adding more variants easier.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
 scripts/checkpatch.pl | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

diff --git a/scripts/checkpatch.pl b/scripts/checkpatch.pl
index 2b3c228..0245bbe 100755
--- a/scripts/checkpatch.pl
+++ b/scripts/checkpatch.pl
@@ -5116,7 +5116,14 @@ sub process {
 			}
 		}
 # check for memory barriers without a comment.
-		if ($line =~ /\b(mb|rmb|wmb|read_barrier_depends|smp_mb|smp_rmb|smp_wmb|smp_read_barrier_depends)\(/) {
+
+		my @barriers = ('mb', 'rmb', 'wmb', 'read_barrier_depends');
+		my @smp_barriers = ('smp_store_release', 'smp_load_acquire', 'smp_store_mb');
+
+		@smp_barriers = (@smp_barriers, map {"smp_" . $_} @barriers);
+		my $all_barriers = join('|', (@barriers, @smp_barriers));
+
+		if ($line =~ /\b($all_barriers)\(/) {
 			if (!ctx_has_comment($first_line, $linenr)) {
 				WARN("MEMORY_BARRIER",
 				     "memory barrier without comment\n" . $herecurr);
-- 
MST

Re: [PATCH 1/3] checkpatch.pl: add missing memory barriers

[Joe Perches]

On Mon, 2016-01-04 at 13:36 +0200, Michael S. Tsirkin wrote:
> SMP-only barriers were missing in checkpatch.pl
> 
> Refactor code slightly to make adding more variants easier.
> 
> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
> ---
>  scripts/checkpatch.pl | 9 ++++++++-
>  1 file changed, 8 insertions(+), 1 deletion(-)
> 
> diff --git a/scripts/checkpatch.pl b/scripts/checkpatch.pl
> index 2b3c228..0245bbe 100755
> --- a/scripts/checkpatch.pl
> +++ b/scripts/checkpatch.pl
> @@ -5116,7 +5116,14 @@ sub process {
>  			}
>  		}
>  # check for memory barriers without a comment.
> -		if ($line =~ /\b(mb|rmb|wmb|read_barrier_depends|smp_mb|smp_rmb|smp_wmb|smp_read_barrier_depends)\(/) {
> +
> +		my @barriers = ('mb', 'rmb', 'wmb', 'read_barrier_depends');
> +		my @smp_barriers = ('smp_store_release', 'smp_load_acquire', 'smp_store_mb');
> +
> +		@smp_barriers = (@smp_barriers, map {"smp_" . $_} @barriers);

I think using map, which so far checkpatch doesn't use,
makes smp_barriers harder to understand and it'd be
better to enumerate them.

> +		my $all_barriers = join('|', (@barriers, @smp_barriers));
> +
> +		if ($line =~ /\b($all_barriers)\(/) {

It would be better to use /\b$all_barriers\s*\(/
as there's no reason for the capture and there
could be a space between the function and the
open parenthesis.


>  			if (!ctx_has_comment($first_line, $linenr)) {
>  				WARN("MEMORY_BARRIER",
>  				     "memory barrier without comment\n" . $herecurr);

Re: [PATCH 1/3] checkpatch.pl: add missing memory barriers

[Russell King - ARM Linux]

On Mon, Jan 04, 2016 at 08:07:40AM -0800, Joe Perches wrote:
> On Mon, 2016-01-04 at 13:36 +0200, Michael S. Tsirkin wrote:
> > +		my $all_barriers = join('|', (@barriers, @smp_barriers));
> > +
> > +		if ($line =~ /\b($all_barriers)\(/) {
> 
> It would be better to use /\b$all_barriers\s*\(/
> as there's no reason for the capture and there
> could be a space between the function and the
> open parenthesis.

I think you mean

	/\b(?:$all_barriers)\s*\(/

as 'all_barriers' will be:

	mb|wmb|rmb|smp_mb|smp_wmb|smp_rmb

and putting that into your suggestion results in:

	/\bmb|wmb|rmb|smp_mb|smp_wmb|smp_rmb\s*\(/

which is clearly wrong - the \b only applies to 'mb' and the \s*\( only
applies to smp_rmb.

-- 
RMK's Patch system: http://www.arm.linux.org.uk/developer/patches/
FTTC broadband for 0.8mile line: currently at 9.6Mbps down 400kbps up
according to speedtest.net.

Re: [PATCH 1/3] checkpatch.pl: add missing memory barriers

[Joe Perches]

On Mon, 2016-01-04 at 16:11 +0000, Russell King - ARM Linux wrote:
> On Mon, Jan 04, 2016 at 08:07:40AM -0800, Joe Perches wrote:
> > On Mon, 2016-01-04 at 13:36 +0200, Michael S. Tsirkin wrote:
> > > +		my $all_barriers = join('|', (@barriers, @smp_barriers));
> > > +
> > > +		if ($line =~ /\b($all_barriers)\(/) {
> > 
> > It would be better to use /\b$all_barriers\s*\(/
> > as there's no reason for the capture and there
> > could be a space between the function and the
> > open parenthesis.
> 
> I think you mean
> 
> 	/\b(?:$all_barriers)\s*\(/
> 
> as 'all_barriers' will be:
> 
> 	mb|wmb|rmb|smp_mb|smp_wmb|smp_rmb
> 
> and putting that into your suggestion results in:
> 
> 	/\bmb|wmb|rmb|smp_mb|smp_wmb|smp_rmb\s*\(/
> 
> which is clearly wrong - the \b only applies to 'mb' and the \s*\( only
> applies to smp_rmb.

right, thanks.

Re: [PATCH 1/3] checkpatch.pl: add missing memory barriers

[Michael S. Tsirkin]

On Mon, Jan 04, 2016 at 08:07:40AM -0800, Joe Perches wrote:
> On Mon, 2016-01-04 at 13:36 +0200, Michael S. Tsirkin wrote:
> > SMP-only barriers were missing in checkpatch.pl
> > 
> > Refactor code slightly to make adding more variants easier.
> > 
> > Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
> > ---
> >  scripts/checkpatch.pl | 9 ++++++++-
> >  1 file changed, 8 insertions(+), 1 deletion(-)
> > 
> > diff --git a/scripts/checkpatch.pl b/scripts/checkpatch.pl
> > index 2b3c228..0245bbe 100755
> > --- a/scripts/checkpatch.pl
> > +++ b/scripts/checkpatch.pl
> > @@ -5116,7 +5116,14 @@ sub process {
> >  			}
> >  		}
> >  # check for memory barriers without a comment.
> > -		if ($line =~ /\b(mb|rmb|wmb|read_barrier_depends|smp_mb|smp_rmb|smp_wmb|smp_read_barrier_depends)\(/) {
> > +
> > +		my @barriers = ('mb', 'rmb', 'wmb', 'read_barrier_depends');
> > +		my @smp_barriers = ('smp_store_release', 'smp_load_acquire', 'smp_store_mb');
> > +
> > +		@smp_barriers = (@smp_barriers, map {"smp_" . $_} @barriers);
> 
> I think using map, which so far checkpatch doesn't use,
> makes smp_barriers harder to understand and it'd be
> better to enumerate them.

Okay - I'll rewrite using foreach.

> > +		my $all_barriers = join('|', (@barriers, @smp_barriers));
> > +
> > +		if ($line =~ /\b($all_barriers)\(/) {
> 
> It would be better to use /\b$all_barriers\s*\(/
> as there's no reason for the capture and there
> could be a space between the function and the
> open parenthesis.

That's the way it was - space before ( will trigger other
warnings. But sure, ok.

> 
> >  			if (!ctx_has_comment($first_line, $linenr)) {
> >  				WARN("MEMORY_BARRIER",
> >  				     "memory barrier without comment\n" . $herecurr);

Re: [PATCH 1/3] checkpatch.pl: add missing memory barriers

[Joe Perches]

On Mon, 2016-01-04 at 22:45 +0200, Michael S. Tsirkin wrote:
> On Mon, Jan 04, 2016 at 08:07:40AM -0800, Joe Perches wrote:
> > On Mon, 2016-01-04 at 13:36 +0200, Michael S. Tsirkin wrote:
> > > SMP-only barriers were missing in checkpatch.pl
> > > 
> > > Refactor code slightly to make adding more variants easier.
> > > 
> > > Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
> > > ---
> > >  scripts/checkpatch.pl | 9 ++++++++-
> > >  1 file changed, 8 insertions(+), 1 deletion(-)
> > > 
> > > diff --git a/scripts/checkpatch.pl b/scripts/checkpatch.pl
> > > index 2b3c228..0245bbe 100755
> > > --- a/scripts/checkpatch.pl
> > > +++ b/scripts/checkpatch.pl
> > > @@ -5116,7 +5116,14 @@ sub process {
> > >  			}
> > >  		}
> > >  # check for memory barriers without a comment.
> > > -		if ($line =~ /\b(mb|rmb|wmb|read_barrier_depends|smp_mb|smp_rmb|smp_wmb|smp_read_barrier_depends)\(/) {
> > > +
> > > +		my @barriers = ('mb', 'rmb', 'wmb', 'read_barrier_depends');
> > > +		my @smp_barriers = ('smp_store_release', 'smp_load_acquire', 'smp_store_mb');
> > > +
> > > +		@smp_barriers = (@smp_barriers, map {"smp_" . $_} @barriers);
> > 
> > I think using map, which so far checkpatch doesn't use,
> > makes smp_barriers harder to understand and it'd be
> > better to enumerate them.
> 
> Okay - I'll rewrite using foreach.
> 

I think using the qr form (like 'our $Attribute' and
a bunch of others) is the general style that should
be used for checkpatch.

Re: [PATCH 1/3] checkpatch.pl: add missing memory barriers

[Michael S. Tsirkin]

On Mon, Jan 04, 2016 at 02:15:50PM -0800, Joe Perches wrote:
> On Mon, 2016-01-04 at 22:45 +0200, Michael S. Tsirkin wrote:
> > On Mon, Jan 04, 2016 at 08:07:40AM -0800, Joe Perches wrote:
> > > On Mon, 2016-01-04 at 13:36 +0200, Michael S. Tsirkin wrote:
> > > > SMP-only barriers were missing in checkpatch.pl
> > > > 
> > > > Refactor code slightly to make adding more variants easier.
> > > > 
> > > > Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
> > > > ---
> > > >  scripts/checkpatch.pl | 9 ++++++++-
> > > >  1 file changed, 8 insertions(+), 1 deletion(-)
> > > > 
> > > > diff --git a/scripts/checkpatch.pl b/scripts/checkpatch.pl
> > > > index 2b3c228..0245bbe 100755
> > > > --- a/scripts/checkpatch.pl
> > > > +++ b/scripts/checkpatch.pl
> > > > @@ -5116,7 +5116,14 @@ sub process {
> > > >  			}
> > > >  		}
> > > >  # check for memory barriers without a comment.
> > > > -		if ($line =~ /\b(mb|rmb|wmb|read_barrier_depends|smp_mb|smp_rmb|smp_wmb|smp_read_barrier_depends)\(/) {
> > > > +
> > > > +		my @barriers = ('mb', 'rmb', 'wmb', 'read_barrier_depends');
> > > > +		my @smp_barriers = ('smp_store_release', 'smp_load_acquire', 'smp_store_mb');
> > > > +
> > > > +		@smp_barriers = (@smp_barriers, map {"smp_" . $_} @barriers);
> > > 
> > > I think using map, which so far checkpatch doesn't use,
> > > makes smp_barriers harder to understand and it'd be
> > > better to enumerate them.
> > 
> > Okay - I'll rewrite using foreach.
> > 
> 
> I think using the qr form (like 'our $Attribute' and
> a bunch of others) is the general style that should
> be used for checkpatch.

Thanks - that's what I did in the new version (included in
v3 of the arch cleanup patchset now).

[PATCH 2/3] checkpatch: check for __smp outside barrier.h

[Michael S. Tsirkin]

Introduction of __smp barriers cleans up a bunch of duplicate code, but
it gives people an additional handle onto a "new" set of barriers - just
because they're prefixed with __* unfortunately doesn't stop anyone from
using it (as happened with other arch stuff before.)

Add a checkpatch test so it will trigger a warning.

Reported-by: Russell King <linux@arm.linux.org.uk>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
 scripts/checkpatch.pl | 11 +++++++++++
 1 file changed, 11 insertions(+)

diff --git a/scripts/checkpatch.pl b/scripts/checkpatch.pl
index 0245bbe..e3f9ad9 100755
--- a/scripts/checkpatch.pl
+++ b/scripts/checkpatch.pl
@@ -5130,6 +5130,17 @@ sub process {
 			}
 		}
 
+		my @underscore_smp_barriers = map {"__" . $_} @smp_barriers;
+		my $underscore_all_barriers = join('|', @underscore_smp_barriers);
+
+		if ($realfile !~ m@^include/asm-generic/@ &&
+		    $realfile !~ m@/barrier\.h$@ &&
+		    $line =~ m/\b($underscore_all_barriers)\(/ &&
+		    $line !~ m/^.\s*\#\s*define\s+($underscore_all_barriers)\(/) {
+			WARN("MEMORY_BARRIER",
+			     "__smp memory barriers shouldn't be used outside barrier.h and asm-generic\n" . $herecurr);
+		}
+
 # check for waitqueue_active without a comment.
 		if ($line =~ /\bwaitqueue_active\s*\(/) {
 			if (!ctx_has_comment($first_line, $linenr)) {
-- 
MST

[PATCH 3/3] checkpatch: add virt barriers

[Michael S. Tsirkin]

Add virt_ barriers to list of barriers to check for
presence of a comment.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
 scripts/checkpatch.pl | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/scripts/checkpatch.pl b/scripts/checkpatch.pl
index e3f9ad9..5fb6ef7 100755
--- a/scripts/checkpatch.pl
+++ b/scripts/checkpatch.pl
@@ -5121,7 +5121,8 @@ sub process {
 		my @smp_barriers = ('smp_store_release', 'smp_load_acquire', 'smp_store_mb');
 
 		@smp_barriers = (@smp_barriers, map {"smp_" . $_} @barriers);
-		my $all_barriers = join('|', (@barriers, @smp_barriers));
+		my @virt_barriers = map {my $l = $_; $l =~ s/smp_/virt_/; $l} @smp_barriers;
+		my $all_barriers = join('|', (@barriers, @smp_barriers, @virt_barriers));
 
 		if ($line =~ /\b($all_barriers)\(/) {
 			if (!ctx_has_comment($first_line, $linenr)) {
-- 
MST

Re: [PATCH 3/3] checkpatch: add virt barriers

[Joe Perches]

On Mon, 2016-01-04 at 13:37 +0200, Michael S. Tsirkin wrote:
> Add virt_ barriers to list of barriers to check for
> presence of a comment.

Are these virt_ barriers used anywhere?

I see some virtio_ barrier like uses.

Re: [PATCH 3/3] checkpatch: add virt barriers

[Michael S. Tsirkin]

On Mon, Jan 04, 2016 at 08:47:53AM -0800, Joe Perches wrote:
> On Mon, 2016-01-04 at 13:37 +0200, Michael S. Tsirkin wrote:
> > Add virt_ barriers to list of barriers to check for
> > presence of a comment.
> 
> Are these virt_ barriers used anywhere?
> 
> I see some virtio_ barrier like uses.

They will be :) They are added and used by patchset
	        arch: barrier cleanup + barriers for virt

See
http://article.gmane.org/gmane.linux.kernel.virtualization/26555


-- 
MST

Re: [PATCH 3/3] checkpatch: add virt barriers

[Joe Perches]

On Mon, 2016-01-04 at 23:07 +0200, Michael S. Tsirkin wrote:
> On Mon, Jan 04, 2016 at 08:47:53AM -0800, Joe Perches wrote:
> > On Mon, 2016-01-04 at 13:37 +0200, Michael S. Tsirkin wrote:
> > > Add virt_ barriers to list of barriers to check for
> > > presence of a comment.
> > 
> > Are these virt_ barriers used anywhere?
> > 
> > I see some virtio_ barrier like uses.
> 
> They will be :) They are added and used by patchset
> 	        arch: barrier cleanup + barriers for virt
> 
> See
> http://article.gmane.org/gmane.linux.kernel.virtualization/26555

Ah, OK, thanks.

Are the virtio_ barriers going away?
If not, maybe those should be added too.

Re: [PATCH 3/3] checkpatch: add virt barriers

[Michael S. Tsirkin]

On Mon, Jan 04, 2016 at 02:11:37PM -0800, Joe Perches wrote:
> On Mon, 2016-01-04 at 23:07 +0200, Michael S. Tsirkin wrote:
> > On Mon, Jan 04, 2016 at 08:47:53AM -0800, Joe Perches wrote:
> > > On Mon, 2016-01-04 at 13:37 +0200, Michael S. Tsirkin wrote:
> > > > Add virt_ barriers to list of barriers to check for
> > > > presence of a comment.
> > > 
> > > Are these virt_ barriers used anywhere?
> > > 
> > > I see some virtio_ barrier like uses.
> > 
> > They will be :) They are added and used by patchset
> > 	        arch: barrier cleanup + barriers for virt
> > 
> > See
> > http://article.gmane.org/gmane.linux.kernel.virtualization/26555
> 
> Ah, OK, thanks.
> 
> Are the virtio_ barriers going away?
> If not, maybe those should be added too.

I don't mind. I'll queue a patch like that.


-- 
MST