[PATCHv2 0/2] device-mapper: use dynamic debug

[PATCHv2 0/2] device-mapper: use dynamic debug

[Hannes Reinecke]

Hi all,

here's an updated version of my earlier patch to switch to dynamic
debug instead of using a compile-time option.
I've modified the initial patch as suggested by Mike, and also fixed
up stray newlines which had crept in with some debugging messages
in dm-zoned.

Hannes Reinecke (2):
  device-mapper: use dynamic debug instead of compile-time config option
  dm-zoned: remove spurious newlines from debugging messages

 drivers/md/dm-zoned-reclaim.c | 4 ++--
 drivers/md/dm-zoned-target.c  | 4 ++--
 include/linux/device-mapper.h | 5 ++---
 3 files changed, 6 insertions(+), 7 deletions(-)

-- 
2.16.4

[PATCH 1/2] device-mapper: use dynamic debug instead of compile-time config option

[Hannes Reinecke]

Switch to use dynamic debug to avoid having recompile the kernel
just to enable debugging messages.

Signed-off-by: Hannes Reinecke <hare@suse.de>
---
 include/linux/device-mapper.h | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/include/linux/device-mapper.h b/include/linux/device-mapper.h
index af48d9da3916..4694e1bb4196 100644
--- a/include/linux/device-mapper.h
+++ b/include/linux/device-mapper.h
@@ -557,12 +557,11 @@ void *dm_vcalloc(unsigned long nmemb, unsigned long elem_size);
 #define DMINFO(fmt, ...) pr_info(DM_FMT(fmt), ##__VA_ARGS__)
 #define DMINFO_LIMIT(fmt, ...) pr_info_ratelimited(DM_FMT(fmt), ##__VA_ARGS__)
 
+#define DMDEBUG_LIMIT(fmt, ...) pr_debug_ratelimited(DM_FMT(fmt), ##__VA_ARGS__)
 #ifdef CONFIG_DM_DEBUG
 #define DMDEBUG(fmt, ...) printk(KERN_DEBUG DM_FMT(fmt), ##__VA_ARGS__)
-#define DMDEBUG_LIMIT(fmt, ...) pr_debug_ratelimited(DM_FMT(fmt), ##__VA_ARGS__)
 #else
-#define DMDEBUG(fmt, ...) no_printk(fmt, ##__VA_ARGS__)
-#define DMDEBUG_LIMIT(fmt, ...) no_printk(fmt, ##__VA_ARGS__)
+#define DMDEBUG(fmt, ...) pr_debug(DM_FMT(fmt), ##__VA_ARGS__)
 #endif
 
 #define DMEMIT(x...) sz += ((sz >= maxlen) ? \
-- 
2.16.4

[PATCH 2/2] dm-zoned: remove spurious newlines from debugging messages

[Hannes Reinecke]

DMDEBUG will already add a newline to the logging messages, so we
shouldn't be adding it to the message itself.

Signed-off-by: Hannes Reinecke <hare@suse.de>
---
 drivers/md/dm-zoned-reclaim.c | 4 ++--
 drivers/md/dm-zoned-target.c  | 4 ++--
 2 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/md/dm-zoned-reclaim.c b/drivers/md/dm-zoned-reclaim.c
index 39ea0d5d4706..4bfa61540b9c 100644
--- a/drivers/md/dm-zoned-reclaim.c
+++ b/drivers/md/dm-zoned-reclaim.c
@@ -405,7 +405,7 @@ static int dmz_do_reclaim(struct dmz_reclaim *zrc)
 
 	ret = dmz_flush_metadata(zrc->metadata);
 	if (ret) {
-		DMDEBUG("(%s): Metadata flush for zone %u failed, err %d\n",
+		DMDEBUG("(%s): Metadata flush for zone %u failed, err %d",
 			dmz_metadata_label(zmd), rzone->id, ret);
 		return ret;
 	}
@@ -493,7 +493,7 @@ static void dmz_reclaim_work(struct work_struct *work)
 
 	ret = dmz_do_reclaim(zrc);
 	if (ret) {
-		DMDEBUG("(%s): Reclaim error %d\n",
+		DMDEBUG("(%s): Reclaim error %d",
 			dmz_metadata_label(zmd), ret);
 		if (!dmz_check_dev(zmd))
 			return;
diff --git a/drivers/md/dm-zoned-target.c b/drivers/md/dm-zoned-target.c
index ea43f6892ced..a3d572da70ad 100644
--- a/drivers/md/dm-zoned-target.c
+++ b/drivers/md/dm-zoned-target.c
@@ -515,7 +515,7 @@ static void dmz_flush_work(struct work_struct *work)
 	/* Flush dirty metadata blocks */
 	ret = dmz_flush_metadata(dmz->metadata);
 	if (ret)
-		DMDEBUG("(%s): Metadata flush failed, rc=%d\n",
+		DMDEBUG("(%s): Metadata flush failed, rc=%d",
 			dmz_metadata_label(dmz->metadata), ret);
 
 	/* Process queued flush requests */
@@ -679,7 +679,7 @@ static int dmz_map(struct dm_target *ti, struct bio *bio)
 	/* Now ready to handle this BIO */
 	ret = dmz_queue_chunk_work(dmz, bio);
 	if (ret) {
-		DMDEBUG("(%s): BIO op %d, can't process chunk %llu, err %i\n",
+		DMDEBUG("(%s): BIO op %d, can't process chunk %llu, err %i",
 			dmz_metadata_label(zmd),
 			bio_op(bio), (u64)dmz_bio_chunk(zmd, bio),
 			ret);
-- 
2.16.4

Re: [PATCH 1/2] device-mapper: use dynamic debug instead of compile-time config option

[Damien Le Moal]

On 2020/05/14 15:09, Hannes Reinecke wrote:
> Switch to use dynamic debug to avoid having recompile the kernel
> just to enable debugging messages.
> 
> Signed-off-by: Hannes Reinecke <hare@suse.de>
> ---
>  include/linux/device-mapper.h | 5 ++---
>  1 file changed, 2 insertions(+), 3 deletions(-)
> 
> diff --git a/include/linux/device-mapper.h b/include/linux/device-mapper.h
> index af48d9da3916..4694e1bb4196 100644
> --- a/include/linux/device-mapper.h
> +++ b/include/linux/device-mapper.h
> @@ -557,12 +557,11 @@ void *dm_vcalloc(unsigned long nmemb, unsigned long elem_size);
>  #define DMINFO(fmt, ...) pr_info(DM_FMT(fmt), ##__VA_ARGS__)
>  #define DMINFO_LIMIT(fmt, ...) pr_info_ratelimited(DM_FMT(fmt), ##__VA_ARGS__)
>  
> +#define DMDEBUG_LIMIT(fmt, ...) pr_debug_ratelimited(DM_FMT(fmt), ##__VA_ARGS__)

Why do you move this one out of the #ifdef CONFIG_DM_DEBUG scope ?

>  #ifdef CONFIG_DM_DEBUG
>  #define DMDEBUG(fmt, ...) printk(KERN_DEBUG DM_FMT(fmt), ##__VA_ARGS__)
> -#define DMDEBUG_LIMIT(fmt, ...) pr_debug_ratelimited(DM_FMT(fmt), ##__VA_ARGS__)
>  #else
> -#define DMDEBUG(fmt, ...) no_printk(fmt, ##__VA_ARGS__)
> -#define DMDEBUG_LIMIT(fmt, ...) no_printk(fmt, ##__VA_ARGS__)
> +#define DMDEBUG(fmt, ...) pr_debug(DM_FMT(fmt), ##__VA_ARGS__)
>  #endif
>  
>  #define DMEMIT(x...) sz += ((sz >= maxlen) ? \
> 


-- 
Damien Le Moal
Western Digital Research

Re: [PATCH 2/2] dm-zoned: remove spurious newlines from debugging messages

[Damien Le Moal]

On 2020/05/14 15:09, Hannes Reinecke wrote:
> DMDEBUG will already add a newline to the logging messages, so we
> shouldn't be adding it to the message itself.
> 
> Signed-off-by: Hannes Reinecke <hare@suse.de>
> ---
>  drivers/md/dm-zoned-reclaim.c | 4 ++--
>  drivers/md/dm-zoned-target.c  | 4 ++--
>  2 files changed, 4 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/md/dm-zoned-reclaim.c b/drivers/md/dm-zoned-reclaim.c
> index 39ea0d5d4706..4bfa61540b9c 100644
> --- a/drivers/md/dm-zoned-reclaim.c
> +++ b/drivers/md/dm-zoned-reclaim.c
> @@ -405,7 +405,7 @@ static int dmz_do_reclaim(struct dmz_reclaim *zrc)
>  
>  	ret = dmz_flush_metadata(zrc->metadata);
>  	if (ret) {
> -		DMDEBUG("(%s): Metadata flush for zone %u failed, err %d\n",
> +		DMDEBUG("(%s): Metadata flush for zone %u failed, err %d",
>  			dmz_metadata_label(zmd), rzone->id, ret);
>  		return ret;
>  	}
> @@ -493,7 +493,7 @@ static void dmz_reclaim_work(struct work_struct *work)
>  
>  	ret = dmz_do_reclaim(zrc);
>  	if (ret) {
> -		DMDEBUG("(%s): Reclaim error %d\n",
> +		DMDEBUG("(%s): Reclaim error %d",
>  			dmz_metadata_label(zmd), ret);
>  		if (!dmz_check_dev(zmd))
>  			return;
> diff --git a/drivers/md/dm-zoned-target.c b/drivers/md/dm-zoned-target.c
> index ea43f6892ced..a3d572da70ad 100644
> --- a/drivers/md/dm-zoned-target.c
> +++ b/drivers/md/dm-zoned-target.c
> @@ -515,7 +515,7 @@ static void dmz_flush_work(struct work_struct *work)
>  	/* Flush dirty metadata blocks */
>  	ret = dmz_flush_metadata(dmz->metadata);
>  	if (ret)
> -		DMDEBUG("(%s): Metadata flush failed, rc=%d\n",
> +		DMDEBUG("(%s): Metadata flush failed, rc=%d",
>  			dmz_metadata_label(dmz->metadata), ret);
>  
>  	/* Process queued flush requests */
> @@ -679,7 +679,7 @@ static int dmz_map(struct dm_target *ti, struct bio *bio)
>  	/* Now ready to handle this BIO */
>  	ret = dmz_queue_chunk_work(dmz, bio);
>  	if (ret) {
> -		DMDEBUG("(%s): BIO op %d, can't process chunk %llu, err %i\n",
> +		DMDEBUG("(%s): BIO op %d, can't process chunk %llu, err %i",
>  			dmz_metadata_label(zmd),
>  			bio_op(bio), (u64)dmz_bio_chunk(zmd, bio),
>  			ret);
> 

Reviewed-by: Damien Le Moal <damien.lemoal@wdc.com>

-- 
Damien Le Moal
Western Digital Research

Re: [PATCH 1/2] device-mapper: use dynamic debug instead of compile-time config option

[Mike Snitzer]

On Thu, May 14 2020 at  3:53am -0400,
Damien Le Moal <Damien.LeMoal@wdc.com> wrote:

> On 2020/05/14 15:09, Hannes Reinecke wrote:
> > Switch to use dynamic debug to avoid having recompile the kernel
> > just to enable debugging messages.
> > 
> > Signed-off-by: Hannes Reinecke <hare@suse.de>
> > ---
> >  include/linux/device-mapper.h | 5 ++---
> >  1 file changed, 2 insertions(+), 3 deletions(-)
> > 
> > diff --git a/include/linux/device-mapper.h b/include/linux/device-mapper.h
> > index af48d9da3916..4694e1bb4196 100644
> > --- a/include/linux/device-mapper.h
> > +++ b/include/linux/device-mapper.h
> > @@ -557,12 +557,11 @@ void *dm_vcalloc(unsigned long nmemb, unsigned long elem_size);
> >  #define DMINFO(fmt, ...) pr_info(DM_FMT(fmt), ##__VA_ARGS__)
> >  #define DMINFO_LIMIT(fmt, ...) pr_info_ratelimited(DM_FMT(fmt), ##__VA_ARGS__)
> >  
> > +#define DMDEBUG_LIMIT(fmt, ...) pr_debug_ratelimited(DM_FMT(fmt), ##__VA_ARGS__)
> 
> Why do you move this one out of the #ifdef CONFIG_DM_DEBUG scope ?

I'd imagine because it already uses dynamic debugging and is useful even
if CONFIG_DM_DEBUG isn't set.  Makes sense to me.  I'll add a note about
it in the commit header though.

Mike

Re: [PATCH 1/2] device-mapper: use dynamic debug instead of compile-time config option

[Mike Snitzer]

On Thu, May 14 2020 at  2:09am -0400,
Hannes Reinecke <hare@suse.de> wrote:

> Switch to use dynamic debug to avoid having recompile the kernel
> just to enable debugging messages.
> 
> Signed-off-by: Hannes Reinecke <hare@suse.de>

FYI, I decided to always use pr_debug* rather than preserve using printk
if CONFIG_DM_DEBUG is enabled.  More consistent and less surprising for
a developer (to have to reason through how to enable debugging).

May cause initial surprise if debugging messages disappear despite
CONFIG_DM_DEBUG being enabled... but pretty sure DM developers can cope
(and/or help non-developers) as needed.

Patches were staged for 5.8, I also added an additional related cleanup:
https://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm.git/log/?h=dm-5.8

Thanks,
Mike