Re: [PATCH v2 2/3] mm: Update file times when inodes are written after mmaped writes

2013-01-03 Thread Jan Kara
On Thu 03-01-13 09:49:37, Andy Lutomirski wrote: > On Mon, Dec 31, 2012 at 8:11 AM, Jan Kara wrote: > > On Sat 22-12-12 00:43:30, Andy Lutomirski wrote: > >> On Sat, Dec 22, 2012 at 12:29 AM, Christoph Hellwig > >> wrote: > >> > NAK, we went through great trouble to get rid of the nasty layering

Re: [PATCH v2 2/3] mm: Update file times when inodes are written after mmaped writes

2013-01-03 Thread Andy Lutomirski
On Mon, Dec 31, 2012 at 8:11 AM, Jan Kara wrote: > On Sat 22-12-12 00:43:30, Andy Lutomirski wrote: >> On Sat, Dec 22, 2012 at 12:29 AM, Christoph Hellwig >> wrote: >> > NAK, we went through great trouble to get rid of the nasty layering >> > violation where the VM called file_update_time direct

Re: [PATCH v2 2/3] mm: Update file times when inodes are written after mmaped writes

2012-12-31 Thread Jan Kara
On Sat 22-12-12 00:43:30, Andy Lutomirski wrote: > On Sat, Dec 22, 2012 at 12:29 AM, Christoph Hellwig > wrote: > > NAK, we went through great trouble to get rid of the nasty layering > > violation where the VM called file_update_time directly just a short > > while ago, reintroducing that is a m

Re: [PATCH v2 2/3] mm: Update file times when inodes are written after mmaped writes

2012-12-24 Thread Zheng Liu
On Fri, Dec 21, 2012 at 01:28:27PM -0800, Andy Lutomirski wrote: > The onus is currently on filesystems to call file_update_time > somewhere in the page_mkwrite path. This is unfortunate for three > reasons: > > 1. page_mkwrite on a locked page should be fast. ext4, for example, >often sleep

Re: [PATCH v2 2/3] mm: Update file times when inodes are written after mmaped writes

2012-12-22 Thread Andy Lutomirski
On Sat, Dec 22, 2012 at 12:29 AM, Christoph Hellwig wrote: > NAK, we went through great trouble to get rid of the nasty layering > violation where the VM called file_update_time directly just a short > while ago, reintroducing that is a massive step back. > > Make sure whatever "solution" for your

Re: [PATCH v2 2/3] mm: Update file times when inodes are written after mmaped writes

2012-12-22 Thread Christoph Hellwig
NAK, we went through great trouble to get rid of the nasty layering violation where the VM called file_update_time directly just a short while ago, reintroducing that is a massive step back. Make sure whatever "solution" for your problem you come up with keeps the file update in the filesystem or

[PATCH v2 2/3] mm: Update file times when inodes are written after mmaped writes

2012-12-21 Thread Andy Lutomirski
The onus is currently on filesystems to call file_update_time somewhere in the page_mkwrite path. This is unfortunate for three reasons: 1. page_mkwrite on a locked page should be fast. ext4, for example, often sleeps while dirtying inodes. (This could be considered a fixable problem with