Skip to contentSkip to conversation

s3c: Fix CONFIG_MTD_NAND_S3C2410_CLKSTOP

PATCH New Christian Pellegrin 2 participants 4 messages Mar 30, 2010 – May 6, 2010

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

PATCH 1/1 s3c: Fix CONFIG_MTD_NAND_S3C2410_CLKSTOP

Option CONFIG_MTD_NAND_S3C2410_CLKSTOP was broken by higher MTD layers calling select_chip with -1 twice in a row. This patch fixes the problem by making sure that clk_disable is not called too many times.

---
 drivers/mtd/nand/s3c2410.c |   10 ++++++++--
 1 files changed, 8 insertions(+), 2 deletions(-)
drivers/mtd/nand/s3c2410.c +8 -2
index fa6e9c7..9adecc6 100644
@@ -37,6 +37,7 @@
 #include <linux/slab.h>
 #include <linux/clk.h>
 #include <linux/cpufreq.h>
+#include <linux/bitops.h>
 
 #include <linux/mtd/mtd.h>
 #include <linux/mtd/nand.h>
@@ -135,6 +136,9 @@ struct s3c2410_nand_info {
 #ifdef CONFIG_CPU_FREQ
 	struct notifier_block	freq_transition;
 #endif
+#ifdef CONFIG_MTD_NAND_S3C2410_CLKSTOP
+	unsigned long                   clk_running;
+#endif
 };
 
 /* conversion functions */
@@ -333,7 +337,8 @@ static void s3c2410_nand_select_chip(struct mtd_info *mtd, int chip)
 	nmtd = this->priv;
 	info = nmtd->info;
 
-	if (chip != -1 && allow_clk_stop(info))
+	if (chip != -1 && allow_clk_stop(info) &&
+	    !test_and_set_bit(0, &info->clk_running))
 		clk_enable(info->clk);
 
 	cur = readl(info->sel_reg);
@@ -356,7 +361,8 @@ static void s3c2410_nand_select_chip(struct mtd_info *mtd, int chip)
 
 	writel(cur, info->sel_reg);
 
-	if (chip == -1 && allow_clk_stop(info))
+	if (chip == -1 && allow_clk_stop(info) &&
+	    test_and_clear_bit(0, &info->clk_running))
 		clk_disable(info->clk);
 }
 
Signature · 1 line
1.5.6.5
#1 lore raw

PATCH 1/1 s3c: Fix CONFIG_MTD_NAND_S3C2410_CLKSTOP

Quoted 32 lines · Christian Pellegrin

Option CONFIG_MTD_NAND_S3C2410_CLKSTOP was broken by higher MTD layers calling select_chip with -1 twice in a row. This patch fixes the problem by making sure that clk_disable is not called too many times.

---
 drivers/mtd/nand/s3c2410.c |   10 ++++++++--
 1 files changed, 8 insertions(+), 2 deletions(-)
drivers/mtd/nand/s3c2410.c +4 -0
index fa6e9c7..9adecc6 100644
@@ -37,6 +37,7 @@
 #include <linux/slab.h>
 #include <linux/clk.h>
 #include <linux/cpufreq.h>
+#include <linux/bitops.h>
 
 #include <linux/mtd/mtd.h>
 #include <linux/mtd/nand.h>
@@ -135,6 +136,9 @@ struct s3c2410_nand_info {
 #ifdef CONFIG_CPU_FREQ
 	struct notifier_block	freq_transition;
 #endif
+#ifdef CONFIG_MTD_NAND_S3C2410_CLKSTOP
+	unsigned long                   clk_running;
+#endif
 };
 
 /* conversion functions */

drivers/mtd/nand/s3c2410.c static void s3c2410_nand_select_chip(struct mtd_info *mtd, int chip) · line 342

Quoted 7 lines
drivers/mtd/nand/s3c2410.c +2 -1
@@ -333,7 +337,8 @@ static void s3c2410_nand_select_chip(struct mtd_info *mtd, int chip)
 	nmtd = this->priv;
 	info = nmtd->info;
 
-	if (chip != -1 && allow_clk_stop(info))
+	if (chip != -1 && allow_clk_stop(info) &&
+	    !test_and_set_bit(0, &info->clk_running))
 		clk_enable(info->clk);

please let me know why you feel the need to use test_and_set_bit on a variable that is single use.

Quoted 18 lines
drivers/mtd/nand/s3c2410.c +2 -1
 	cur = readl(info->sel_reg);
@@ -356,7 +361,8 @@ static void s3c2410_nand_select_chip(struct mtd_info *mtd, int chip)
 
 	writel(cur, info->sel_reg);
 
-	if (chip == -1 && allow_clk_stop(info))
+	if (chip == -1 && allow_clk_stop(info) &&
+	    test_and_clear_bit(0, &info->clk_running))
 		clk_disable(info->clk);
 }
 
Signature · 6 lines
1.5.6.5

--
To unsubscribe from this list: send the line "unsubscribe linux-samsung-soc" in
the body of a message to majordomo at vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Signature · 5 lines
-- 
Ben

Q:      What's a light-year?
A:      One-third less calories than a regular year.
#2 lore raw

PATCH 1/1 s3c: Fix CONFIG_MTD_NAND_S3C2410_CLKSTOP

Quoted 4 lines · Ben Dooks

please let me know why you feel the need to use test_and_set_bit on a variable that is single use.

As I mentioned in the description the select function is not called in a balanced way by upper layers, so there is a need to count it in the driver (otherwise you end by calling clk_disable too many times if I remember well). Perhaps the atomic operations are not needed but I'm not sure about this so I preferred to take the cautionary approach (I haven't studied if calls to select are serialized in the upper MTD layer or even the block layer). Feel free to use a simple integer, but I confirm that something like this patch is needed in 2.6.33 and 2.6.34 if you want clock stopping enabled.

Thanks,

Signature · 5 lines
Christian Pellegrin, see http://www.evolware.org/chri/
"Real Programmers don't play tennis, or any other sport which requires
you to change clothes. Mountain climbing is OK, and Real Programmers
wear their climbing boots to work in case a mountain should suddenly
spring up in the middle of the computer room."
#3 lore raw

PATCH 1/1 s3c: Fix CONFIG_MTD_NAND_S3C2410_CLKSTOP

Quoted 6 lines · christian pellegrin
Quoted 4 lines · Ben Dooks

please let me know why you feel the need to use test_and_set_bit on a variable that is single use.

Quoted 9 lines

As I mentioned in the description the select function is not called in a balanced way by upper layers, so there is a need to count it in the driver (otherwise you end by calling clk_disable too many times if I remember well). Perhaps the atomic operations are not needed but I'm not sure about this so I preferred to take the cautionary approach (I haven't studied if calls to select are serialized in the upper MTD layer or even the block layer). Feel free to use a simple integer, but I confirm that something like this patch is needed in 2.6.33 and 2.6.34 if you want clock stopping enabled.

I don't belive that atomic ops are needed, IIRC the mtd layer should sort that out. As such test_and_set_bit is hardly a count?

Quoted 12 lines

Thanks,

Signature · 9 lines
Christian Pellegrin, see http://www.evolware.org/chri/
"Real Programmers don't play tennis, or any other sport which requires
you to change clothes. Mountain climbing is OK, and Real Programmers
wear their climbing boots to work in case a mountain should suddenly
spring up in the middle of the computer room."
--
To unsubscribe from this list: send the line "unsubscribe linux-samsung-soc" in
the body of a message to majordomo at vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Signature · 5 lines
-- 
Ben

Q:      What's a light-year?
A:      One-third less calories than a regular year.