Re: [PATCH] AFS fixes and improvements

2009-07-20 Thread Pavel Roskin
On Mon, 2009-07-20 at 20:56 +0200, Vladimir 'phcoder' Serbinenko wrote: > I'm ok with it in condition that pre-release freeze period won't last > long. Now Marco Gerards is just back from vacation and I propose to > wait for his statements rather than speaking about virtual release OK -- Regard

Re: [PATCH] AFS fixes and improvements

2009-07-20 Thread Vladimir 'phcoder' Serbinenko
On Mon, Jul 20, 2009 at 8:34 PM, Pavel Roskin wrote: > On Mon, 2009-07-20 at 11:36 +0200, Vladimir 'phcoder' Serbinenko wrote: >> On Sun, Jul 19, 2009 at 10:47 PM, Pavel Roskin wrote: >> > I think we should release 1.97 first.  Sorry, I know that it's bad to >> > sit on a pile of code, but in my op

Re: [PATCH] AFS fixes and improvements

2009-07-20 Thread Pavel Roskin
On Mon, 2009-07-20 at 11:36 +0200, Vladimir 'phcoder' Serbinenko wrote: > On Sun, Jul 19, 2009 at 10:47 PM, Pavel Roskin wrote: > > I think we should release 1.97 first. Sorry, I know that it's bad to > > sit on a pile of code, but in my opinion, the nested partition patch is > > too intrusive at

Re: [PATCH] AFS fixes and improvements

2009-07-20 Thread Vladimir 'phcoder' Serbinenko
On Sun, Jul 19, 2009 at 10:47 PM, Pavel Roskin wrote: > On Sun, 2009-07-19 at 22:40 +0200, Vladimir 'phcoder' Serbinenko wrote: >> > I prefer not to fix minor memory leaks by huge patches.  I have a patch, >> > I just wanted to look at the possibility to use grub_kzalloc() and >> > allocating data

Re: [PATCH] AFS fixes and improvements

2009-07-19 Thread Pavel Roskin
On Sun, 2009-07-19 at 22:40 +0200, Vladimir 'phcoder' Serbinenko wrote: > > I prefer not to fix minor memory leaks by huge patches. I have a patch, > > I just wanted to look at the possibility to use grub_kzalloc() and > > allocating data as part of the partition. > Well nested partition patch cha

Re: [PATCH] AFS fixes and improvements

2009-07-19 Thread Vladimir 'phcoder' Serbinenko
> I prefer not to fix minor memory leaks by huge patches.  I have a patch, > I just wanted to look at the possibility to use grub_kzalloc() and > allocating data as part of the partition. Well nested partition patch changes the way how partitions are handled and fix for memory leak probably wouldn'

Re: [PATCH] AFS fixes and improvements

2009-07-19 Thread Pavel Roskin
On Sun, 2009-07-19 at 22:23 +0200, Vladimir 'phcoder' Serbinenko wrote: > Yes, I did it too and valgrind only issued some warnings about my -O3 > libraries and found a memory leak caused by pc_partition_map_parse. I know. > This function anyway will be removed with my nested partition patch > (B

Re: [PATCH] AFS fixes and improvements

2009-07-19 Thread Vladimir 'phcoder' Serbinenko
On Sun, Jul 19, 2009 at 9:56 PM, Pavel Roskin wrote: > On Sun, 2009-07-19 at 17:34 +0200, Vladimir 'phcoder' Serbinenko wrote: >> >> It's better to split fixes from the new features. >> Attached patches >> >> >> >> I don't have afs images around.  It would be great if you test all new >> >> functio

Re: [PATCH] AFS fixes and improvements

2009-07-19 Thread Vladimir 'phcoder' Serbinenko
On Sun, Jul 19, 2009 at 10:10 PM, Pavel Roskin wrote: > On Sun, 2009-07-19 at 15:20 +0200, Vladimir 'phcoder' Serbinenko wrote: > >> > -  if ((! dir->inode.stream.size) || >> > +  if ((dir->inode.stream.size == 0) || >> > >> > The later is marginally better, but it would be easier to review your >>

Re: [PATCH] AFS fixes and improvements

2009-07-19 Thread Pavel Roskin
On Sun, 2009-07-19 at 15:20 +0200, Vladimir 'phcoder' Serbinenko wrote: > > - if ((! dir->inode.stream.size) || > > + if ((dir->inode.stream.size == 0) || > > > > The later is marginally better, but it would be easier to review your > > patches if you don't include such changes. > It's not a sty

Re: [PATCH] AFS fixes and improvements

2009-07-19 Thread Pavel Roskin
On Sun, 2009-07-19 at 17:34 +0200, Vladimir 'phcoder' Serbinenko wrote: > >> It's better to split fixes from the new features. > Attached patches > >> > >> I don't have afs images around. It would be great if you test all new > >> functionality with valgrind. It's very good at finding mistakes in

Re: [PATCH] AFS fixes and improvements

2009-07-19 Thread Vladimir 'phcoder' Serbinenko
>> It's better to split fixes from the new features. Attached patches >> >> I don't have afs images around.  It would be great if you test all new >> functionality with valgrind.  It's very good at finding mistakes in the >> code. >> > I will do. I use the publically-available image from > http://w

Re: [PATCH] AFS fixes and improvements

2009-07-19 Thread Vladimir 'phcoder' Serbinenko
On Sun, Jul 19, 2009 at 7:11 AM, Pavel Roskin wrote: > Quoting Vladimir 'phcoder' Serbinenko : > >> Update: added symlink support >> >> On Fri, Jul 17, 2009 at 9:30 PM, Vladimir 'phcoder' >> Serbinenko wrote: >>> >>> Hello. Currently I'm coding BFS (filesystem of BeOS and Haiku) which >>> is simila

Re: [PATCH] AFS fixes and improvements

2009-07-18 Thread Pavel Roskin
Quoting Vladimir 'phcoder' Serbinenko : Update: added symlink support On Fri, Jul 17, 2009 at 9:30 PM, Vladimir 'phcoder' Serbinenko wrote: Hello. Currently I'm coding BFS (filesystem of BeOS and Haiku) which is similar to AFS. So I took the later as codebase. I found some bugs and incompleten

Re: [PATCH] AFS fixes and improvements

2009-07-17 Thread Vladimir 'phcoder' Serbinenko
Update: added symlink support On Fri, Jul 17, 2009 at 9:30 PM, Vladimir 'phcoder' Serbinenko wrote: > Hello. Currently I'm coding BFS (filesystem of BeOS and Haiku) which > is similar to AFS. So I took the later as codebase. I found some bugs > and incompletenesses in it. Here is the fix. Tested u

[PATCH] AFS fixes and improvements

2009-07-17 Thread Vladimir 'phcoder' Serbinenko
Hello. Currently I'm coding BFS (filesystem of BeOS and Haiku) which is similar to AFS. So I took the later as codebase. I found some bugs and incompletenesses in it. Here is the fix. Tested using grub-fstest on virtual machine image downloaded from Syllable website and then successfully booted fro