Skip to contentSkip to conversation

Add ide-disk led-trigger to ATA subsystem

PATCH In review Carsten Schlote 5 participants 5 messages Jan 10, 2009 – Jan 18, 2009 libata

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 Add ide-disk led-trigger to ATA subsystem

- Adds a call to ide-disk LED trigger to libata-sff.c
- Changed Kconfig to allow for ide-disk LED trigger when ATA drivers are enabled.

Open Issue:
- There might be more suitable place to call the trigger function. Only devices using libata-sff.c trigger LEDs right now.

---
 drivers/ata/libata-sff.c |    3 +++
 drivers/leds/Kconfig     |    2 +-
 2 files changed, 4 insertions(+), 1 deletion(-)

Index: .kernel/drivers/leds/Kconfig
===================================================================
--- .kernel.orig/drivers/leds/Kconfig 2009-01-07 22:13:40.000000000
+0100
+++ .kernel/drivers/leds/Kconfig 2009-01-07 22:14:46.000000000 +0100
@@ -185,7 +185,7 @@

config LEDS_TRIGGER_IDE_DISK
bool "LED IDE Disk Trigger"
- depends on LEDS_TRIGGERS && BLK_DEV_IDEDISK
+ depends on LEDS_TRIGGERS && BLK_DEV_IDEDISK || ATA
help
This allows LEDs to be controlled by IDE disk activity. If unsure, say Y.
Index: .kernel/drivers/ata/libata-sff.c
===================================================================
--- .kernel.orig/drivers/ata/libata-sff.c 2009-01-07 22:14:09.000000000
+0100
+++ .kernel/drivers/ata/libata-sff.c 2009-01-07 22:14:46.000000000 +0100
@@ -36,6 +36,7 @@
#include <linux/pci.h>
#include <linux/libata.h>
#include <linux/highmem.h>
+#include <linux/leds.h>

#include "libata.h"

@@ -647,6 +648,8 @@
*/
void ata_sff_exec_command(struct ata_port *ap, const struct
ata_taskfile *tf)
{
+ ledtrig_ide_activity();
+
DPRINTK("ata%u: cmd 0x%X\n", ap->print_id, tf->command);

iowrite8(tf->command, ap->ioaddr.command_addr);

#1 lore raw

(cc linux-ide)

Quoted 52 lines · On Sat, 10 Jan 2009 16:08:27 +0100 Carsten Schlote <[email protected]>…

- Adds a call to ide-disk LED trigger to libata-sff.c
- Changed Kconfig to allow for ide-disk LED trigger when ATA drivers are enabled.

Open Issue:
- There might be more suitable place to call the trigger function. Only devices using libata-sff.c trigger LEDs right now.

---
 drivers/ata/libata-sff.c |    3 +++
 drivers/leds/Kconfig     |    2 +-
 2 files changed, 4 insertions(+), 1 deletion(-)

Index: .kernel/drivers/leds/Kconfig
===================================================================
--- .kernel.orig/drivers/leds/Kconfig 2009-01-07 22:13:40.000000000
+0100
+++ .kernel/drivers/leds/Kconfig 2009-01-07 22:14:46.000000000 +0100

(quoted diff) +1 -1
@@ -185,7 +185,7 @@
 
 config LEDS_TRIGGER_IDE_DISK
 	bool "LED IDE Disk Trigger"
-	depends on LEDS_TRIGGERS && BLK_DEV_IDEDISK
+	depends on LEDS_TRIGGERS && BLK_DEV_IDEDISK || ATA
 	help
 	  This allows LEDs to be controlled by IDE disk activity.
 	  If unsure, say Y.

Index: .kernel/drivers/ata/libata-sff.c
===================================================================
--- .kernel.orig/drivers/ata/libata-sff.c 2009-01-07 22:14:09.000000000
+0100
+++ .kernel/drivers/ata/libata-sff.c 2009-01-07 22:14:46.000000000 +0100

(quoted diff) +3 -0
@@ -36,6 +36,7 @@
 #include <linux/pci.h>
 #include <linux/libata.h>
 #include <linux/highmem.h>
+#include <linux/leds.h>
 
 #include "libata.h"
 
@@ -647,6 +648,8 @@
  */
 void ata_sff_exec_command(struct ata_port *ap, const struct
 ata_taskfile *tf)
 {
+	ledtrig_ide_activity();
+
 	DPRINTK("ata%u: cmd 0x%X\n", ap->print_id, tf->command);
 

iowrite8(tf->command, ap->ioaddr.command_addr);

#2 lore raw
In reply to Carsten Schlote
Quoted 20 lines · Carsten Schlote

- Adds a call to ide-disk LED trigger to libata-sff.c
- Changed Kconfig to allow for ide-disk LED trigger when ATA drivers are enabled.

Open Issue:
- There might be more suitable place to call the trigger function. Only devices using libata-sff.c trigger LEDs right now.

---
 drivers/ata/libata-sff.c |    3 +++
 drivers/leds/Kconfig     |    2 +-
 2 files changed, 4 insertions(+), 1 deletion(-)

Index: .kernel/drivers/leds/Kconfig
===================================================================
--- .kernel.orig/drivers/leds/Kconfig 2009-01-07 22:13:40.000000000
+0100
+++ .kernel/drivers/leds/Kconfig 2009-01-07 22:14:46.000000000 +0100

Review line 188

Quoted 5 lines
(quoted diff) +1 -1
@@ -185,7 +185,7 @@
 
 config LEDS_TRIGGER_IDE_DISK
 	bool "LED IDE Disk Trigger"
-	depends on LEDS_TRIGGERS && BLK_DEV_IDEDISK
+	depends on LEDS_TRIGGERS && BLK_DEV_IDEDISK || ATA

()s might be nice?

Quoted 11 lines
(quoted diff) +2 -0
@@ -647,6 +648,8 @@
  */
 void ata_sff_exec_command(struct ata_port *ap, const struct
 ata_taskfile *tf)
 {
+	ledtrig_ide_activity();
+
 	DPRINTK("ata%u: cmd 0x%X\n", ap->print_id, tf->command);
 

iowrite8(tf->command, ap->ioaddr.command_addr);

Is there a more generic place so that we won't have to modify all the libata drivers?
Pavel

Signature · 2 lines
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
#3 lore raw
Quoted 8 lines · Pavel Machek
Quoted 7 lines · Carsten Schlote

void ata_sff_exec_command(struct ata_port *ap, const struct
ata_taskfile *tf)
{

(quoted diff) +2 -0
+	ledtrig_ide_activity();
+
 	DPRINTK("ata%u: cmd 0x%X\n", ap->print_id, tf->command);
Quoted 5 lines
Quoted 2 lines

iowrite8(tf->command, ap->ioaddr.command_addr);

Is there a more generic place so that we won't have to modify all the libata drivers?

One would think you could do this at the block layer...

Jeff

#4 lore raw
Quoted 9 lines · On Sun, 18 Jan 2009 10:50:30 -0500 Jeff Garzik <[email protected]> wrote:
Quoted 8 lines · Pavel Machek
Quoted 7 lines · Carsten Schlote

void ata_sff_exec_command(struct ata_port *ap, const struct
ata_taskfile *tf)
{

(quoted diff) +2 -0
+	ledtrig_ide_activity();
+
 	DPRINTK("ata%u: cmd 0x%X\n", ap->print_id, tf->command);
Quoted 8 lines
Quoted 5 lines
Quoted 2 lines

iowrite8(tf->command, ap->ioaddr.command_addr);

Is there a more generic place so that we won't have to modify all the libata drivers?

One would think you could do this at the block layer...

ata_sff_exec_command is also on a timing critical path so the wrong place to be touching other stuff. It's mid command issue.