Skip to contentSkip to conversation

x86/resctrl: Use seq_putc() in two functions

PATCH In review Markus Elfring 4 participants 6 messages Jul 13, 2024 – Jul 17, 2024

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 x86/resctrl: Use seq_putc() in two functions

From: Markus Elfring <[email protected]>
Date: Sat, 13 Jul 2024 10:43:14 +0200

Single characters should be put into a sequence.
Thus use the corresponding function “seq_putc”.

This issue was transformed by using the Coccinelle software.

---
 arch/x86/kernel/cpu/resctrl/ctrlmondata.c | 4 ++--
 arch/x86/kernel/cpu/resctrl/rdtgroup.c    | 4 ++--
 2 files changed, 4 insertions(+), 4 deletions(-)
arch/x86/kernel/cpu/resctrl/ctrlmondata.c +2 -2
index 50fa1fe9a073..dce68f19c29d 100644
@@ -452,7 +452,7 @@ static void show_doms(struct seq_file *s, struct resctrl_schema *schema, int clo
 	seq_printf(s, "%*s:", max_name_width, schema->name);
 	list_for_each_entry(dom, &r->ctrl_domains, hdr.list) {
 		if (sep)
-			seq_puts(s, ";");
+			seq_putc(s, ';');
 
 		if (is_mba_sc(r))
 			ctrl_val = dom->mbps_val[closid];
@@ -464,7 +464,7 @@ static void show_doms(struct seq_file *s, struct resctrl_schema *schema, int clo
 			   ctrl_val);
 		sep = true;
 	}
-	seq_puts(s, "\n");
+	seq_putc(s, '\n');
 }
 
 int rdtgroup_schemata_show(struct kernfs_open_file *of,
arch/x86/kernel/cpu/resctrl/rdtgroup.c +2 -2
index d7163b764c62..d4e216073e7a 100644
@@ -1605,7 +1605,7 @@ static int mbm_config_show(struct seq_file *s, struct rdt_resource *r, u32 evtid
 
 	list_for_each_entry(dom, &r->mon_domains, hdr.list) {
 		if (sep)
-			seq_puts(s, ";");
+			seq_putc(s, ';');
 
 		memset(&mon_info, 0, sizeof(struct mon_config_info));
 		mon_info.evtid = evtid;
@@ -1614,7 +1614,7 @@ static int mbm_config_show(struct seq_file *s, struct rdt_resource *r, u32 evtid
 		seq_printf(s, "%d=0x%02x", dom->hdr.id, mon_info.mon_config);
 		sep = true;
 	}
-	seq_puts(s, "\n");
+	seq_putc(s, '\n');
 
 	mutex_unlock(&rdtgroup_mutex);
 	cpus_read_unlock();

--
2.45.2

maintainer
#1 lore raw

Hi Markus,

Quoted 7 lines · Markus Elfring

From: Markus Elfring <[email protected]>
Date: Sat, 13 Jul 2024 10:43:14 +0200

Single characters should be put into a sequence.
Thus use the corresponding function “seq_putc”.

This issue was transformed by using the Coccinelle software.

Could you please point me to how you accomplished this? I tried on the latest upstream as well as the latest resctrl staged changes in tip's x86/cache but the needed coccinelle scripts do not seem to be available. I tried by running the commands below. It appeared to run many coccinelle scripts but there was not one related to this change. Is this a new coccinelle script outside of the kernel or still on its way upstream?

make C=2 CHECK=scripts/coccicheck MODE=report arch/x86/kernel/cpu/resctrl/rdtgroup.o make C=2 CHECK=scripts/coccicheck MODE=report arch/x86/kernel/cpu/resctrl/ctrlmondata.o

Quoted 52 lines
---
  arch/x86/kernel/cpu/resctrl/ctrlmondata.c | 4 ++--
  arch/x86/kernel/cpu/resctrl/rdtgroup.c    | 4 ++--
  2 files changed, 4 insertions(+), 4 deletions(-)
arch/x86/kernel/cpu/resctrl/ctrlmondata.c +2 -2
index 50fa1fe9a073..dce68f19c29d 100644
@@ -452,7 +452,7 @@ static void show_doms(struct seq_file *s, struct resctrl_schema *schema, int clo
  	seq_printf(s, "%*s:", max_name_width, schema->name);
  	list_for_each_entry(dom, &r->ctrl_domains, hdr.list) {
  		if (sep)
-			seq_puts(s, ";");
+			seq_putc(s, ';');
 
  		if (is_mba_sc(r))
  			ctrl_val = dom->mbps_val[closid];
@@ -464,7 +464,7 @@ static void show_doms(struct seq_file *s, struct resctrl_schema *schema, int clo
  			   ctrl_val);
  		sep = true;
  	}
-	seq_puts(s, "\n");
+	seq_putc(s, '\n');
  }
 
  int rdtgroup_schemata_show(struct kernfs_open_file *of,
arch/x86/kernel/cpu/resctrl/rdtgroup.c +2 -3
index d7163b764c62..d4e216073e7a 100644
@@ -1605,7 +1605,7 @@ static int mbm_config_show(struct seq_file *s, struct rdt_resource *r, u32 evtid
 
  	list_for_each_entry(dom, &r->mon_domains, hdr.list) {
  		if (sep)
-			seq_puts(s, ";");
+			seq_putc(s, ';');
 
  		memset(&mon_info, 0, sizeof(struct mon_config_info));
  		mon_info.evtid = evtid;
@@ -1614,7 +1614,7 @@ static int mbm_config_show(struct seq_file *s, struct rdt_resource *r, u32 evtid
  		seq_printf(s, "%d=0x%02x", dom->hdr.id, mon_info.mon_config);
  		sep = true;
  	}
-	seq_puts(s, "\n");
+	seq_putc(s, '\n');
 
  	mutex_unlock(&rdtgroup_mutex);
  	cpus_read_unlock();
--
Quoted 2 lines

2.45.2

Could you please highlight the benefit of this change? Looking at seq_puts() implementation, thanks to [1], it seems to me these seq_puts() calls will result in seq_putc() anyway?

Reinette

[1] https://lore.kernel.org/all/a8589bffe4830dafcb9111e22acf06603fea7132.1713781332.git.christophe.jaillet@wanadoo.fr/

#2 lore raw
Quoted 19 lines · Reinette Chatre

Hi Markus,

Quoted 7 lines · Markus Elfring

From: Markus Elfring <[email protected]>
Date: Sat, 13 Jul 2024 10:43:14 +0200

Single characters should be put into a sequence.
Thus use the corresponding function “seq_putc”.

This issue was transformed by using the Coccinelle software.

Could you please point me to how you accomplished this? I tried on the latest upstream as well as the latest resctrl staged changes in tip's x86/cache but the needed coccinelle scripts do not seem to be available. I tried by running the commands below. It appeared to run many coccinelle scripts but there was not one related to this change. Is this a new coccinelle script outside of the kernel or still on its way upstream?

Quoted 2 lines

make C=2 CHECK=scripts/coccicheck MODE=report arch/x86/kernel/cpu/resctrl/rdtgroup.o make C=2 CHECK=scripts/coccicheck MODE=report arch/x86/kernel/cpu/resctrl/ctrlmondata.o

Just to give some context to this: Markus usually uses extra (out-of-tree) coccinelle scripts to change code to his liking, sometimes these are good changes, sometimes they are meaningless.

See for example this:

https://lore.kernel.org/all/[email protected]/

Some well known kernel contributors also have auto-replies (although in a different context) specially for this contributor:

https://lore.kernel.org/all/2024071537-schematic-envoy-4272@gregkh/

Cheers,
Chris

#3 lore raw
Quoted 2 lines

Just to give some context to this: Markus usually uses extra (out-of-tree) coccinelle scripts to change code to his liking,

I occasionally dare to present special change possibilities.

Quoted 1 line

sometimes these are good changes,

These can be nice.

Quoted 1 line

sometimes they are meaningless.

Further development ideas can trigger corresponding communication challenges.

Regards,
Markus

#4 lore raw
In reply to Reinette Chatre

Le 16/07/2024 à 23:43, Reinette Chatre a écrit :

Quoted 9 lines

Hi Markus,

Quoted 6 lines · Markus Elfring

From: Markus Elfring <[email protected]>
Date: Sat, 13 Jul 2024 10:43:14 +0200

Single characters should be put into a sequence.
Thus use the corresponding function “seq_putc”.

...

Quoted 2 lines

Could you please highlight the benefit of this change? Looking at seq_puts() implementation, thanks to [1], it seems to me these seq_puts() calls will result in seq_putc() anyway?

Hi,

in this case, there is no benefit and the generated code should be exactly the same.

CJ

Quoted 6 lines

Reinette

[1] https://lore.kernel.org/all/a8589bffe4830dafcb9111e22acf06603fea7132.1713781332.git.christophe.jaillet@wanadoo.fr/

#5 lore raw
In reply to Reinette Chatre
Quoted 6 lines
Quoted 4 lines

Single characters should be put into a sequence.
Thus use the corresponding function “seq_putc”.

This issue was transformed by using the Coccinelle software.

Could you please point me to how you accomplished this?

…

Maybe.

Quoted 1 line

Is this a new coccinelle script outside of the kernel

I constructed small scripts for the semantic patch language according to the presented code refactoring a while ago.

Quoted 1 line

or still on its way upstream?

I am curious if such a transformation approach can eventually be integrated for the coccicheck tool.

Quoted 2 lines

Could you please highlight the benefit of this change? Looking at seq_puts() implementation, thanks to [1], it seems to me these seq_puts() calls will result in seq_putc() anyway?

If the appropriate function call would be directly used, extra optimisation efforts can be avoided by the compiler.

Regards,
Markus