On Wed, Apr 11, 2007 at 07:10:36PM -0700, Haifeng He wrote:
> I have a question about trap handling in UML kernel, particularly in the
> skas0 mode.
> My understanding is that, UML will fork a kernel thread for each user
> process
UML will fork a host process for each UML address space.
> and
>
Hi, All
I have a question about trap handling in UML kernel, particularly in the
skas0 mode.
My understanding is that, UML will fork a kernel thread for each user
process and
this thread will go to a loop (in userspace()) and wait for signal from user
process using
ptrace. If a signal occurs, it
On Wed, Apr 11, 2007 at 01:05:16PM -0400, Gerb Stralko wrote:
> Is this wrong? you declare io_reg as a pointer but then you pass him
> to os_write_file_k by reference of a pointer? Should you just do this
> instead:
Intentional, I'm writing pointers, not structures, down the pipe.
Tx for the re
On 4/11/07, Jeff Dike <[EMAIL PROTECTED]> wrote:
> Instead of writing entire structures between UML and the I/O thread,
> we send pointers. This cuts down on the amount of data being copied
> and possibly allows more requests to be pending between the two.
>
> - n = os_write
These are 2.6.22 material. Two of them are improvements to the block
driver, one causes UML to dump core on a panic, and the other is a bug
fix.
Jeff
--
Work email - jdike at linux dot intel dot com
---
Dump core after a panic. This will provide better debugging
information than is currently available.
Signed-off-by: Jeff Dike <[EMAIL PROTECTED]>
--
arch/um/include/os.h |1 +
arch/um/kernel/um_arch.c |2 +-
arch/um/os-Linux/util.c |6 ++
3 files changed, 8 insertions(+), 1
Code running on the initial UML stack can't receive or process signals
since current must be valid when IRQs are handled, and there is no
current for this stack.
So, instead of using UML_LONGJMP and UML_SETJMP, which are careful
to save and restore signal state, and, as a side-effect, handle any
d
Instead of writing entire structures between UML and the I/O thread,
we send pointers. This cuts down on the amount of data being copied
and possibly allows more requests to be pending between the two.
This requires that the requests be kmalloced and freed instead of
living on the stack.
Signed-
Send as many I/O requests to the I/O thread as possible, even though
it will still only handle one at a time. This provides an opportunity
to reduce latency by starting one request before the previous one has
been finished in the driver.
Request handling is somewhat modernized by requesting sg pi
