| Message ID | 1664a8ea72b34db4bef860004166875ab70898ee.1433442778.git.joe@perches.com (mailing list archive) |
|---|---|
| State | Rejected, archived |
| Headers | show |
On Tue, 2015-06-09 at 01:35 +0200, Rafael J. Wysocki wrote: > On Thursday, June 04, 2015 11:36:44 AM Joe Perches wrote: > > Add the ability to see how long it takes to sync the filesystems > > via the printk time mechanism. > > > > Start to standardize the printk "PM: doing something...done" > > messages on two separate lines. > > Well, it would be good to say what problem this is attempting to fix. > > And while I understand the underlying concern, there is a merit in keeping > each of these messages in one line (if everything goes well), so I'm wondering > what about printing each of them in one go after the operation with a tail > depending on the result? Like > > printk(KERN_INFO "PM: Syncing filesystems ... done\n"); > > on success or > > printk(KERN_INFO "PM: Syncing filesystems ... failed\n"); > > on failure? Maybe. I believe there are multi-second delays possible when syncing the filesystems on things like usb memory sticks. I think the dmesg line count here isn't particularly important and there's some small value in consistently presenting timing information via dmesg timestamps when using 2 lines. There's also some small value in patch 6/6 and the consistent prefixing of "PM: " on all these messages. -- To unsubscribe from this list: send the line "unsubscribe linux-pm" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Thursday, June 04, 2015 11:36:44 AM Joe Perches wrote: > Add the ability to see how long it takes to sync the filesystems > via the printk time mechanism. > > Start to standardize the printk "PM: doing something...done" > messages on two separate lines. Well, it would be good to say what problem this is attempting to fix. And while I understand the underlying concern, there is a merit in keeping each of these messages in one line (if everything goes well), so I'm wondering what about printing each of them in one go after the operation with a tail depending on the result? Like printk(KERN_INFO "PM: Syncing filesystems ... done\n"); on success or printk(KERN_INFO "PM: Syncing filesystems ... failed\n"); on failure? > Signed-off-by: Joe Perches <joe@perches.com> > --- > kernel/power/hibernate.c | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-) > > diff --git a/kernel/power/hibernate.c b/kernel/power/hibernate.c > index 2329daa..2466d78 100644 > --- a/kernel/power/hibernate.c > +++ b/kernel/power/hibernate.c > @@ -663,9 +663,9 @@ int hibernate(void) > if (error) > goto Exit; > > - printk(KERN_INFO "PM: Syncing filesystems ... "); > + printk(KERN_INFO "PM: Syncing filesystems ...\n"); > sys_sync(); > - printk("done.\n"); > + printk(KERN_INFO "PM: Syncing filesystems: done\n"); > > error = freeze_processes(); > if (error) >
On Mon 2015-06-08 16:23:24, Joe Perches wrote: > On Tue, 2015-06-09 at 01:35 +0200, Rafael J. Wysocki wrote: > > On Thursday, June 04, 2015 11:36:44 AM Joe Perches wrote: > > > Add the ability to see how long it takes to sync the filesystems > > > via the printk time mechanism. > > > > > > Start to standardize the printk "PM: doing something...done" > > > messages on two separate lines. > > > > Well, it would be good to say what problem this is attempting to fix. > > > > And while I understand the underlying concern, there is a merit in keeping > > each of these messages in one line (if everything goes well), so I'm wondering > > what about printing each of them in one go after the operation with a tail > > depending on the result? Like > > > > printk(KERN_INFO "PM: Syncing filesystems ... done\n"); > > > > on success or > > > > printk(KERN_INFO "PM: Syncing filesystems ... failed\n"); > > > > on failure? > > Maybe. > > I believe there are multi-second delays possible when > syncing the filesystems on things like usb memory sticks. > > I think the dmesg line count here isn't particularly important > and there's some small value in consistently presenting timing > information via dmesg timestamps when using 2 lines. Hibernation is little special here, because that's what user actually sees, and because message interleaving is extremely unlikely there. Unless you have a real problem with this, just let it be. Pavel
diff --git a/kernel/power/hibernate.c b/kernel/power/hibernate.c index 2329daa..2466d78 100644 --- a/kernel/power/hibernate.c +++ b/kernel/power/hibernate.c @@ -663,9 +663,9 @@ int hibernate(void) if (error) goto Exit; - printk(KERN_INFO "PM: Syncing filesystems ... "); + printk(KERN_INFO "PM: Syncing filesystems ...\n"); sys_sync(); - printk("done.\n"); + printk(KERN_INFO "PM: Syncing filesystems: done\n"); error = freeze_processes(); if (error)
Add the ability to see how long it takes to sync the filesystems via the printk time mechanism. Start to standardize the printk "PM: doing something...done" messages on two separate lines. Signed-off-by: Joe Perches <joe@perches.com> --- kernel/power/hibernate.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)