Re: [PATCH] fs/devfs/base.c

2001-06-06 Thread Pavel Machek
Hi! > > It's trivial to calculate for DAGs -- directed acyclic graphs. It's > > when the "acyclic" constraint is violated that you have problems! > > It may well be that interrupt stacks are a win anyway. If we can get the kernel > struct out of the stack pages (which would fix some very unplea

Re: [PATCH] fs/devfs/base.c

2001-06-05 Thread Andrew Morton
Alan Cox wrote: > > > It's trivial to calculate for DAGs -- directed acyclic graphs. It's > > when the "acyclic" constraint is violated that you have problems! > > It may well be that interrupt stacks are a win anyway. If we can get the kernel > struct out of the stack pages (which would fix so

Re: [PATCH] fs/devfs/base.c

2001-06-05 Thread Ingo Molnar
On Mon, 4 Jun 2001, Alan Cox wrote: > > - the kernel stack is 4kB, and _nobody_ has the right to eat up a > >noticeable portion of it. It doesn't matter if you "know" your caller > > Umm Linus on what platform - its 8K or more on all that I can think of it's 8K-sizeof(struct task_struct).

Re: [PATCH] fs/devfs/base.c

2001-06-04 Thread H. Peter Anvin
Followup to: <[EMAIL PROTECTED]> By author:Bill Pringlemeir <[EMAIL PROTECTED]> In newsgroup: linux.dev.kernel > > There was a discussion on comp.arch.embedded about bounded stack use. > It is fairly easy to calculate the stack usage for call trees, but > much more difficult for `DAGs'. Ie,

Kernel Stack usage [was: [PATCH] fs/devfs/base.c]

2001-06-04 Thread Bill Pringlemeir
> Bill Pringlemeir <[EMAIL PROTECTED]> writes: > There was a discussion on comp.arch.embedded about bounded stack > use. It is fairly easy to calculate the stack usage for call > trees, but much more difficult for `DAGs'. Ie, a recursive > functions etc. I don't know about the poli

Re: [PATCH] fs/devfs/base.c

2001-06-04 Thread Bill Pringlemeir
> "Richard" == Richard Gooch <[EMAIL PROTECTED]> writes: [snip] > I guess we should ask the question as to what's an > acceptable usage. Theoretically, any amount could pose a > problem, but that's hardly a useful position to work There was a discussion on comp.arch.embedded about bounded

Maximum size of automatic allocation? (was: [PATCH] fs/devfs/base.c)

2001-05-28 Thread Daniel Phillips
On Sunday 27 May 2001 15:21, Richard Gooch wrote: > Akash Jain writes: > > in fs/devfs/base.c, > > the struct devfsd_notify_struct is approx 1056 bytes, allocating it > > on a stack of 8k seems unreasonable. here we simply move it to the > > heap, i don't think it is a _must_ be on stack type thi

Re: [PATCH] fs/devfs/base.c

2001-05-27 Thread Richard Gooch
Akash Jain writes: > hello, > > in fs/devfs/base.c, > the struct devfsd_notify_struct is approx 1056 bytes, allocating it > on a stack of 8k seems unreasonable. here we simply move it to the > heap, i don't think it is a _must_ be on stack type thing I absolutely don't want this patch applied.

[PATCH] fs/devfs/base.c

2001-05-27 Thread Akash Jain
hello, in fs/devfs/base.c, the struct devfsd_notify_struct is approx 1056 bytes, allocating it on a stack of 8k seems unreasonable. here we simply move it to the heap, i don't think it is a _must_ be on stack type thing -aki- --- fs/devfs/base.c.origTue Apr 17 15:04:10 2001 +++ fs/dev