Skip to contentSkip to conversation

pxa-devel back-merge / cherry pick request

PATCH New Daniel Mack 5 participants 11 messages May 31, 2010

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

Hi Eric,

Is there any plans to merge Linus' tree into your pxa-devel any time soon? I'm asking because I have one patch pending which depends on

a3bcbbee ("pda_power: Add function callbacks for suspend and resume")

upstream. Alternatively, could you cherry-pick this one commit?

Thanks,
Daniel

#1 lore raw
Quoted 8 lines · Daniel Mack

Hi Eric,

Is there any plans to merge Linus' tree into your pxa-devel any time soon? I'm asking because I have one patch pending which depends on

a3bcbbee ("pda_power: Add function callbacks for suspend and resume")

upstream. Alternatively, could you cherry-pick this one commit?

Please don't cherry-pick commits back into branches which will eventually find their way back into mainline.

#2 lore raw
Quoted 6 lines · Russell King - ARM Linux
Quoted 5 lines · Daniel Mack

Hi Eric,

Is there any plans to merge Linus' tree into your pxa-devel any time soon? I'm asking because I have one patch pending which depends on

Quoted 6 lines
Quoted 3 lines

a3bcbbee ("pda_power: Add function callbacks for suspend and resume")

upstream. Alternatively, could you cherry-pick this one commit?

Please don't cherry-pick commits back into branches which will eventually find their way back into mainline.

Hmm, ok just wait for the back-merge then. It's just that one single commit and not an urgent fix.

Just out of interest - why would cherry-picking be frowned upon? Shouldn't git just ignore the commits it already knows when things are pulled?

Thanks,
Daniel

#3 lore raw
Quoted 3 lines · Daniel Mack

Just out of interest - why would cherry-picking be frowned upon? Shouldn't git just ignore the commits it already knows when things are pulled?

No, git is content based. The merge should be a noop merge if the two sides end up being the same but you still end up with two different copies of the change in history.

#4 lore raw
Quoted 9 lines · Mark Brown
Quoted 3 lines · Daniel Mack

Just out of interest - why would cherry-picking be frowned upon? Shouldn't git just ignore the commits it already knows when things are pulled?

No, git is content based. The merge should be a noop merge if the two sides end up being the same but you still end up with two different copies of the change in history.

Ah, the commit won't have the same ID anymore after it is cherry-picked, of course.

That makes sense. Thanks for explaining.

Daniel

#5 lore raw
In reply to Mark Brown
Quoted 9 lines · Mark Brown
Quoted 3 lines · Daniel Mack

Just out of interest - why would cherry-picking be frowned upon? Shouldn't git just ignore the commits it already knows when things are pulled?

No, git is content based. The merge should be a noop merge if the two sides end up being the same but you still end up with two different copies of the change in history.

Sometimes it actually causes merge conflicts if there are changes in the same part of the file as the picked patch touches which aren't also cherry-picked.

It's best to avoid cherry-picking commits into two different git threads and then try to merge them.

#6 lore raw
In reply to Daniel Mack
Quoted 8 lines · Daniel Mack

Hi Eric,

Is there any plans to merge Linus' tree into your pxa-devel any time soon? I'm asking because I have one patch pending which depends on

a3bcbbee ("pda_power: Add function callbacks for suspend and resume")

upstream. Alternatively, could you cherry-pick this one commit?

Even better is for you to fork a branch starting at that particular commit, apply your patch on top, then ask Eric to pull it.

Nicolas

#7 lore raw
In reply to Daniel Mack
Quoted 9 lines · Daniel Mack

Hi Eric,

Is there any plans to merge Linus' tree into your pxa-devel any time soon? I'm asking because I have one patch pending which depends on

?a3bcbbee ("pda_power: Add function callbacks for suspend and resume")

upstream. Alternatively, could you cherry-pick this one commit?

'devel' branch rebased and pushed. Normally I rebase before sending git pull request to Russell.

#8 lore raw

PATCH 1/1 ARM: pxa/raumfeld: check charge state after resume

Use the resume callback of the pda_power supply framework to check for a 'charge finished' event that might have occured during the sleep phase.

---
 arch/arm/mach-pxa/raumfeld.c |   29 +++++++++++++++++++++--------
 1 files changed, 21 insertions(+), 8 deletions(-)
arch/arm/mach-pxa/raumfeld.c +21 -8
index 1c8cd39..7582a12 100644
@@ -750,13 +750,32 @@ static int raumfeld_is_usb_online(void)
 
 static char *raumfeld_power_supplicants[] = { "ds2760-battery.0" };
 
+static void raumfeld_power_signal_charged(void)
+{
+	struct power_supply *psy =
+		power_supply_get_by_name(raumfeld_power_supplicants[0]);
+
+	if (psy)
+		power_supply_set_battery_charged(psy);
+}
+
+static int raumfeld_power_resume(void)
+{
+	/* check if GPIO_CHARGE_DONE went low while we were sleeping */
+	if (!gpio_get_value(GPIO_CHARGE_DONE))
+		raumfeld_power_signal_charged();
+
+	return 0;
+}
+
 static struct pda_power_pdata power_supply_info = {
 	.init			= power_supply_init,
 	.is_ac_online		= raumfeld_is_ac_online,
 	.is_usb_online		= raumfeld_is_usb_online,
 	.exit			= power_supply_exit,
 	.supplied_to		= raumfeld_power_supplicants,
-	.num_supplicants	= ARRAY_SIZE(raumfeld_power_supplicants)
+	.num_supplicants	= ARRAY_SIZE(raumfeld_power_supplicants),
+	.resume			= raumfeld_power_resume,
 };
 
 static struct resource power_supply_resources[] = {
@@ -771,13 +790,7 @@ static struct resource power_supply_resources[] = {
 
 static irqreturn_t charge_done_irq(int irq, void *dev_id)
 {
-	struct power_supply *psy;
-
-	psy = power_supply_get_by_name("ds2760-battery.0");
-
-	if (psy)
-		power_supply_set_battery_charged(psy);
-
+	raumfeld_power_signal_charged();
 	return IRQ_HANDLED;
 }
 
Signature · 1 line
1.7.1
Applied
#9 lore raw

PATCH 1/1 ARM: pxa/raumfeld: check charge state after resume

Quoted 5 lines · Daniel Mack

Use the resume callback of the pda_power supply framework to check for a 'charge finished' event that might have occured during the sleep phase.

Applied.

Quoted 61 lines

---
?arch/arm/mach-pxa/raumfeld.c | ? 29 +++++++++++++++++++++-------- ?1 files changed, 21 insertions(+), 8 deletions(-)

arch/arm/mach-pxa/raumfeld.c +21 -9
index 1c8cd39..7582a12 100644
@@ -750,13 +750,32 @@ static int raumfeld_is_usb_online(void)
 
 ?static char *raumfeld_power_supplicants[] = { "ds2760-battery.0" };
 
+static void raumfeld_power_signal_charged(void)
+{
+ ? ? ? struct power_supply *psy =
+ ? ? ? ? ? ? ? power_supply_get_by_name(raumfeld_power_supplicants[0]);
+
+ ? ? ? if (psy)
+ ? ? ? ? ? ? ? power_supply_set_battery_charged(psy);
+}
+
+static int raumfeld_power_resume(void)
+{
+ ? ? ? /* check if GPIO_CHARGE_DONE went low while we were sleeping */
+ ? ? ? if (!gpio_get_value(GPIO_CHARGE_DONE))
+ ? ? ? ? ? ? ? raumfeld_power_signal_charged();
+
+ ? ? ? return 0;
+}
+
 ?static struct pda_power_pdata power_supply_info = {
 ? ? ? ?.init ? ? ? ? ? ? ? ? ? = power_supply_init,
 ? ? ? ?.is_ac_online ? ? ? ? ? = raumfeld_is_ac_online,
 ? ? ? ?.is_usb_online ? ? ? ? ?= raumfeld_is_usb_online,
 ? ? ? ?.exit ? ? ? ? ? ? ? ? ? = power_supply_exit,
 ? ? ? ?.supplied_to ? ? ? ? ? ?= raumfeld_power_supplicants,
- ? ? ? .num_supplicants ? ? ? ?= ARRAY_SIZE(raumfeld_power_supplicants)
+ ? ? ? .num_supplicants ? ? ? ?= ARRAY_SIZE(raumfeld_power_supplicants),
+ ? ? ? .resume ? ? ? ? ? ? ? ? = raumfeld_power_resume,
 ?};
 
 ?static struct resource power_supply_resources[] = {
@@ -771,13 +790,7 @@ static struct resource power_supply_resources[] = {
 
 ?static irqreturn_t charge_done_irq(int irq, void *dev_id)
 ?{
- ? ? ? struct power_supply *psy;
-
- ? ? ? psy = power_supply_get_by_name("ds2760-battery.0");
-
- ? ? ? if (psy)
- ? ? ? ? ? ? ? power_supply_set_battery_charged(psy);
-
+ ? ? ? raumfeld_power_signal_charged();
 ? ? ? ?return IRQ_HANDLED;
 ?}
 
--

1.7.1

#10 lore raw
In reply to Eric Miao
Quoted 6 lines · Eric Miao
Quoted 5 lines · Daniel Mack

Hi Eric,

Is there any plans to merge Linus' tree into your pxa-devel any time soon? I'm asking because I have one patch pending which depends on

Quoted 7 lines
Quoted 4 lines

?a3bcbbee ("pda_power: Add function callbacks for suspend and resume")

upstream. Alternatively, could you cherry-pick this one commit?

'devel' branch rebased and pushed. Normally I rebase before sending git pull request to Russell.

It's best to avoid rebasing altogether unless there is a good reason to do so.

If you want to make a decision about whether to rebase, one way to do it is to create a test branch/tree, and try merging your selected rebase point in - if it conflicts then you know when the original branch is merged it'll have problems.