Re: [PATCH 1/1] pytest: Use --lazy with umount
On Wed, Apr 05, 2023 at 10:19:39PM -0400, Tom Rini wrote: > Sometimes when doing tests on real hardware we sometimes run in to the > case where some of these mounts haven't been fully flushed. Using the > --lazy option with umount will allow us to continue while letting the OS > handle flushing the data out still. > > Signed-off-by: Tom Rini > Reviewed-by: Simon Glass Applied to u-boot/master, thanks! -- Tom signature.asc Description: PGP signature
Re: [PATCH 1/1] pytest: Use --lazy with umount
Hi Tom, On Sat, 8 Apr 2023 at 07:46, Tom Rini wrote: > > On Sat, Apr 08, 2023 at 06:55:14AM +1200, Simon Glass wrote: > > Hi Tom, > > > > On Sat, 8 Apr 2023 at 02:36, Tom Rini wrote: > > > > > > On Fri, Apr 07, 2023 at 05:31:17PM +1200, Simon Glass wrote: > > > > Hi Tom, > > > > > > > > On Thu, 6 Apr 2023 at 14:19, Tom Rini wrote: > > > > > > > > > > Sometimes when doing tests on real hardware we sometimes run in to the > > > > > case where some of these mounts haven't been fully flushed. Using the > > > > > --lazy option with umount will allow us to continue while letting the > > > > > OS > > > > > handle flushing the data out still. > > > > > > > > > > Signed-off-by: Tom Rini > > > > > --- > > > > > test/py/tests/test_ut.py | 4 ++-- > > > > > 1 file changed, 2 insertions(+), 2 deletions(-) > > > > > > > > Reviewed-by: Simon Glass > > > > > > > > I wonder if these ever actually succeed later on, or do they remain > > > > mounted forever? > > > > > > They complete the unmount with this change done, yes. > > > > That's great to hear. I wonder if we should try a 'sync' beforehand? > > So long as it is short it shouldn't matter, but if it dragged on for a > > minute or two it might conflict with another job. I don't have a good > > understanding of what is going on in the kernel. > > Based on a quick read of the man page, I think --lazy itself is just > what we want here. OK good. Regards, Simon
Re: [PATCH 1/1] pytest: Use --lazy with umount
On Sat, Apr 08, 2023 at 06:55:14AM +1200, Simon Glass wrote: > Hi Tom, > > On Sat, 8 Apr 2023 at 02:36, Tom Rini wrote: > > > > On Fri, Apr 07, 2023 at 05:31:17PM +1200, Simon Glass wrote: > > > Hi Tom, > > > > > > On Thu, 6 Apr 2023 at 14:19, Tom Rini wrote: > > > > > > > > Sometimes when doing tests on real hardware we sometimes run in to the > > > > case where some of these mounts haven't been fully flushed. Using the > > > > --lazy option with umount will allow us to continue while letting the OS > > > > handle flushing the data out still. > > > > > > > > Signed-off-by: Tom Rini > > > > --- > > > > test/py/tests/test_ut.py | 4 ++-- > > > > 1 file changed, 2 insertions(+), 2 deletions(-) > > > > > > Reviewed-by: Simon Glass > > > > > > I wonder if these ever actually succeed later on, or do they remain > > > mounted forever? > > > > They complete the unmount with this change done, yes. > > That's great to hear. I wonder if we should try a 'sync' beforehand? > So long as it is short it shouldn't matter, but if it dragged on for a > minute or two it might conflict with another job. I don't have a good > understanding of what is going on in the kernel. Based on a quick read of the man page, I think --lazy itself is just what we want here. -- Tom signature.asc Description: PGP signature
Re: [PATCH 1/1] pytest: Use --lazy with umount
Hi Tom, On Sat, 8 Apr 2023 at 02:36, Tom Rini wrote: > > On Fri, Apr 07, 2023 at 05:31:17PM +1200, Simon Glass wrote: > > Hi Tom, > > > > On Thu, 6 Apr 2023 at 14:19, Tom Rini wrote: > > > > > > Sometimes when doing tests on real hardware we sometimes run in to the > > > case where some of these mounts haven't been fully flushed. Using the > > > --lazy option with umount will allow us to continue while letting the OS > > > handle flushing the data out still. > > > > > > Signed-off-by: Tom Rini > > > --- > > > test/py/tests/test_ut.py | 4 ++-- > > > 1 file changed, 2 insertions(+), 2 deletions(-) > > > > Reviewed-by: Simon Glass > > > > I wonder if these ever actually succeed later on, or do they remain > > mounted forever? > > They complete the unmount with this change done, yes. That's great to hear. I wonder if we should try a 'sync' beforehand? So long as it is short it shouldn't matter, but if it dragged on for a minute or two it might conflict with another job. I don't have a good understanding of what is going on in the kernel. Regards, Simon
Re: [PATCH 1/1] pytest: Use --lazy with umount
On Fri, Apr 07, 2023 at 05:31:17PM +1200, Simon Glass wrote: > Hi Tom, > > On Thu, 6 Apr 2023 at 14:19, Tom Rini wrote: > > > > Sometimes when doing tests on real hardware we sometimes run in to the > > case where some of these mounts haven't been fully flushed. Using the > > --lazy option with umount will allow us to continue while letting the OS > > handle flushing the data out still. > > > > Signed-off-by: Tom Rini > > --- > > test/py/tests/test_ut.py | 4 ++-- > > 1 file changed, 2 insertions(+), 2 deletions(-) > > Reviewed-by: Simon Glass > > I wonder if these ever actually succeed later on, or do they remain > mounted forever? They complete the unmount with this change done, yes. -- Tom signature.asc Description: PGP signature
Re: [PATCH 1/1] pytest: Use --lazy with umount
Hi Tom, On Thu, 6 Apr 2023 at 14:19, Tom Rini wrote: > > Sometimes when doing tests on real hardware we sometimes run in to the > case where some of these mounts haven't been fully flushed. Using the > --lazy option with umount will allow us to continue while letting the OS > handle flushing the data out still. > > Signed-off-by: Tom Rini > --- > test/py/tests/test_ut.py | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-) Reviewed-by: Simon Glass I wonder if these ever actually succeed later on, or do they remain mounted forever? Regards, SImon