Peter Wemm cc66ebe2a9 Commit a partial lazy thread switch mechanism for i386. it isn't as lazy ... as it could be and can do with some more cleanup. Currently its under options LAZY_SWITCH. What this does is avoid %cr3 reloads for short context switches that do not involve another user process. ie: we can take an interrupt, switch to a kthread and return to the user without explicitly flushing the tlb. However, this isn't as exciting as it could be, the interrupt overhead is still high and too much blocks on Giant still. There are some debug sysctls, for stats and for an on/off switch. The main problem with doing this has been "what if the process that you're running on exits while we're borrowing its address space?" - in this case we use an IPI to give it a kick when we're about to reclaim the pmap. Its not compiled in unless you add the LAZY_SWITCH option. I want to fix a few more things and get some more feedback before turning it on by default. This is NOT a replacement for Bosko's lazy interrupt stuff. This was more meant for the kthread case, while his was for interrupts. Mine helps a little for interrupts, but his helps a lot more. The stats are enabled with options SWTCH_OPTIM_STATS - this has been a pseudo-option for years, I just added a bunch of stuff to it. One non-trivial change was to select a new thread before calling cpu_switch() in the first place. This allows us to catch the silly case of doing a cpu_switch() to the current process. This happens uncomfortably often. This simplifies a bit of the asm code in cpu_switch (no longer have to call choosethread() in the middle). This has been implemented on i386 and (thanks to jake) sparc64. The others will come soon. This is actually seperate to the lazy switch stuff. Glanced at by: jake, jhb		2003-04-02 23:53:30 +00:00
..
alpha	- Define a new md function 'casuptr'. This atomically compares and sets	2003-04-01 00:18:55 +00:00
amd64	Commit a partial lazy thread switch mechanism for i386. it isn't as lazy	2003-04-02 23:53:30 +00:00
arm	- Add vm_paddr_t, a physical address type. This is required for systems	2003-03-25 00:07:06 +00:00
boot	FreeBSD 5.0 has stopped shipping /modules 2.5 years ago. Catch	2003-03-11 12:09:25 +00:00
cam	Use bioq_flush() to drain a bio queue with a specific error code.	2003-04-01 15:06:26 +00:00
coda	- Move p->p_sigmask to td->td_sigmask. Signal masks will be per thread with	2003-03-31 22:49:17 +00:00
compat	Add #include <sys/conf.h> so we don't rely on <sys/disk.h> doing it.	2003-04-01 12:34:47 +00:00
conf	Commit a partial lazy thread switch mechanism for i386. it isn't as lazy	2003-04-02 23:53:30 +00:00
contrib	Use bioq_flush() to drain a bio queue with a specific error code.	2003-04-01 15:06:26 +00:00
crypto	Remove some unnecessary casts.	2003-01-25 22:41:22 +00:00
ddb	Change the process flags P_KSES to be P_THREADED.	2003-02-27 02:05:19 +00:00
dev	Convert the fxp(4) driver to the busdma API.	2003-04-02 16:47:16 +00:00
fs	- smb_td_intr takes a thread as an argument not a proc.	2003-04-01 09:24:12 +00:00
geom	Add handling for cancelled events in the g_call_me() methods.	2003-04-02 21:10:04 +00:00
gnu	- Lock the buf before clearing flags.	2003-03-13 07:07:16 +00:00
i4b	- Add vm_paddr_t, a physical address type. This is required for systems	2003-03-25 00:07:06 +00:00
i386	Commit a partial lazy thread switch mechanism for i386. it isn't as lazy	2003-04-02 23:53:30 +00:00
ia64	Use bioq_flush() to drain a bio queue with a specific error code.	2003-04-01 15:06:26 +00:00
isa	Use bioq_flush() to drain a bio queue with a specific error code.	2003-04-01 15:06:26 +00:00
isofs/cd9660	Rename vfs_stdsync function to vfs_stdnosync which matches more	2003-03-11 22:15:10 +00:00
kern	Commit a partial lazy thread switch mechanism for i386. it isn't as lazy	2003-04-02 23:53:30 +00:00
libkern	Further GC of M_STRING, missed previously.	2003-02-26 01:00:29 +00:00
modules	opt_geom.h is no longer required for this module.	2003-03-29 16:13:11 +00:00
net	o netisr_queue() returns 1 on success and 0 on failure,	2003-03-27 12:52:57 +00:00
netatalk	- Use if_broadcastaddr from struct ifnet rather than relying on	2003-03-21 17:53:16 +00:00
netatm	Update netisr handling; Each SWI now registers its queue, and all queue	2003-03-04 23:19:55 +00:00
netgraph	Include correct opt_* headers for supported address families. Dike out	2003-03-08 16:25:27 +00:00
netinet	Back out support for RFC3514.	2003-04-02 20:14:44 +00:00
netinet6	fixed a mbuf leak when an IP packet from ESP tunnel is redirected	2003-03-29 08:31:28 +00:00
netipsec	add missing copyright notices	2003-03-28 20:28:05 +00:00
netipx	Remove unimplemented IP-in-IPX encapsulation support (options IPTUNNEL).	2003-03-08 06:58:22 +00:00
netkey	Back out M_* changes, per decision of the TRB.	2003-02-19 05:47:46 +00:00
netnatm	Update netisr handling; Each SWI now registers its queue, and all queue	2003-03-04 23:19:55 +00:00
netncp	- Move p->p_sigmask to td->td_sigmask. Signal masks will be per thread with	2003-03-31 22:49:17 +00:00
netsmb	- Move p->p_sigmask to td->td_sigmask. Signal masks will be per thread with	2003-03-31 22:49:17 +00:00
nfs	Back out M_* changes, per decision of the TRB.	2003-02-19 05:47:46 +00:00
nfsclient	Revert change 1.201 (removing mapping of VAPPEND to VWRITE).	2003-03-31 23:26:10 +00:00
nfsserver	- Lock bufs before inspecting their flags.	2003-03-13 07:05:22 +00:00
opencrypto	Gigacommit to improve device-driver source compatibility between	2003-03-03 12:15:54 +00:00
pc98	MFi386: revisions 1.556 and 1.557.	2003-04-02 10:51:06 +00:00
pccard	Gigacommit to improve device-driver source compatibility between	2003-03-03 12:15:54 +00:00
pci	Fix KASSERT syntax error.	2003-04-01 08:57:28 +00:00
posix4	Back out M_* changes, per decision of the TRB.	2003-02-19 05:47:46 +00:00
powerpc	- Define a new md function 'casuptr'. This atomically compares and sets	2003-04-01 00:18:55 +00:00
rpc
security	Trim "trustedbsd_" from the front of the policy module "short names";	2003-03-27 19:26:39 +00:00
sparc64	Commit a partial lazy thread switch mechanism for i386. it isn't as lazy	2003-04-02 23:53:30 +00:00
sys	Commit a partial lazy thread switch mechanism for i386. it isn't as lazy	2003-04-02 23:53:30 +00:00
tools	MFp4:	2003-03-28 06:43:50 +00:00
ufs	Check return value from vmapbuf instead of the function address.	2003-03-27 20:48:34 +00:00
vm	Add a facility allowing processes to inform the VM subsystem they are	2003-03-31 21:09:57 +00:00
Makefile