On Tue, Nov 09, 2010 at 18:19 -0800, Philip Guenther wrote:
> On Tue, Nov 9, 2010 at 10:41 AM, Mike Belopuhov wrote:
> > On Sun, Oct 31, 2010 at 17:10 -0700, Philip Guenther wrote:
> ...
> >> The TSS also contains the pointer to the stack used for double faults.
> >> Previously, this was placed on
> Date: Tue, 9 Nov 2010 18:19:08 -0800
> From: Philip Guenther
>
> >> /* exceptions */
> >> for (x = 0; x < 32; x++) {
> >> ist = (x == 8) ? 1 : 0;
> >> + // ist = (x == 8) ? 1 : (x == 2) ? 2 : 0;
> >
> > why do you need this comment? or is it a reminder to
On Tue, Nov 9, 2010 at 10:41 AM, Mike Belopuhov wrote:
> On Sun, Oct 31, 2010 at 17:10 -0700, Philip Guenther wrote:
...
>> The TSS also contains the pointer to the stack used for double faults.
>> Previously, this was placed one page above each process's PCB. Rather
>> than change that on each c
On Sun, Oct 31, 2010 at 17:10 -0700, Philip Guenther wrote:
> Okay, enough cleanup; time to simplify the TSS handling by moving from
> TSS-per-process to TSS-per-CPU. This eliminates the need to ever change
> GDT entries after startup and the associated ~4k process limit.
>
> Where do we place
Hi Philip,
Tested very well in build kernel, build userland, build xenocara.
No particular difference in time noted. userland about 70 min,
xenocara about 60 min (59m57 real, 32m user, 33m system).
also tested on another build of mine where i saw the compile drive all
4 CPU's to 100% with make -
On Sun, Oct 31, 2010 at 05:10:46PM -0700, Philip Guenther wrote:
> Okay, enough cleanup; time to simplify the TSS handling by moving from
> TSS-per-process to TSS-per-CPU. This eliminates the need to ever change
> GDT entries after startup and the associated ~4k process limit.
>
> Where do we p
Okay, enough cleanup; time to simplify the TSS handling by moving from
TSS-per-process to TSS-per-CPU. This eliminates the need to ever change
GDT entries after startup and the associated ~4k process limit.
Where do we place that TSS? Before, it was part of the PCB; here I've put
the primary
Thank you again to the people who tested the first diff. Here's diff two
in the cleanup of the amd64 low-level segment bits. Now that the LDT is
only used by dead compat code, load the ldt register with the null
selector (disabling use of it), stop reloading it on every context switch,
and bl
On Sat, Oct 09, 2010 at 01:11:05PM -0700, Philip Guenther wrote:
> The diff below is the first step in a clean up of the amd64 low-level
> segment bits. This step is to switch user-space to using code and data
> segments in the global descriptor table (GDT) instead of the local
> descriptor tab
> Date: Sat, 9 Oct 2010 13:11:05 -0700
> From: Philip Guenther
>
> The diff below is the first step in a clean up of the amd64 low-level
> segment bits. This step is to switch user-space to using code and data
> segments in the global descriptor table (GDT) instead of the local
> descriptor t
The diff below is the first step in a clean up of the amd64 low-level
segment bits. This step is to switch user-space to using code and data
segments in the global descriptor table (GDT) instead of the local
descriptor table (LDT) and to eliminate the GDT slots that we don't
actually use.
Aft
11 matches
Mail list logo
