Hello,
The 0.9.9 2.4.10pre4 patch doesn't apply. There are 5 rejects. The
first three are easy to fix, but
the fourth in vmscan.c will be very difficult without being a developer.
The fifth I am not sure about.
This brings up the fact that many of us are using ext3 and have to patch
every time a new kernel
comes out. I really think we need to strong arm Linus or something and
get ext3 into 2.4.11 at
least as expermintal. If that is near impossible then we should really
look into a autopatch
generator like i2c has you just do something like mkpatch /usr/src/linux
| patch -p1 and it auto
generates a valid patch file for that version for you. Which makes it as
simple as running the
command and making a new patch instead of waiting for the developers to
come out with a
new one because of small changes in most cases.
"Nathan G. Grennan" wrote:> The 0.9.9 2.4.10pre4 patch doesn't apply. There are 5 rejects. The > first three are easy to fix, but > the fourth in vmscan.c will be very difficult without being a developer. > The fifth I am not sure about.In the post pre4 pre patches, the last two patches weren't required (or rather, i didnt think they were and things seems to work fine for me without them.> > This brings up the fact that many of us are using ext3 and have to patch > every time a new kernel > comes out. I really think we need to strong arm Linus or something and > get ext3 into 2.4.11 at > least as expermintal. If that is near impossible then we should reallyThat would be nice.> look into a autopatch > generator like i2c has you just do something like mkpatch /usr/src/linux > | patch -p1 and it auto > generates a valid patch file for that version for you. Which makes it as > simple as running the > command and making a new patch instead of waiting for the developers to > come out with a > new one because of small changes in most cases.Dunno if autopatching would work since the kernel is changing alot in regards with VM and VFS stuff. It'd just break ext3. i2c had the same problem a while back when a kernel call was renamed. -- ====================================================================Mohammad A. Haque http://www.haque.net/ mhaque@haque.net "Alcohol and calculus don't mix. Project Lead Don't drink and derive." --Unknown http://wm.themes.org/ batmanppc@themes.org =====================================================================
"Nathan G. Grennan" wrote:> > Hello, > > The 0.9.9 2.4.10pre4 patch doesn't apply.Linus seems to want an ext3 merge soon. I'll put 0.9.10 together and if it's OK with sct we'll try for a 2.4.11-pre[1 or 2] merge. Perhaps.
Hi, On Sun, Sep 23, 2001 at 04:28:46PM -0500, Nathan G. Grennan wrote:> least as expermintal. If that is near impossible then we should really > look into a autopatch > generator like i2c has you just do something like mkpatch /usr/src/linux > | patch -p1 and it auto > generates a valid patch file for that version for you.Trouble is, the entire reason you get conflicts when applying old ext3 patches to new kernels is that the underlying semantics of the VFS and VM in the core kernel are changing rapidly these days. You _can't_ automate a solution to that. In most cases, if the relevant parts of the underlying kernel have not changed, then you can keep on using the same ext3 patch. --Stephen