lk/linux-kernel-test
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

lguest: improve interrupt handling, speed up stream networking

Browse Source 
lguest never checked for pending interrupts when enabling interrupts, and
things still worked.  However, it makes a significant difference to TCP
performance, so it's time we fixed it by introducing a pending_irq flag
and checking it on irq_restore and irq_enable.

These two routines are now too big to patch into the 8/10 bytes
patch space, so we drop that code.

Note: The high latency on interrupt delivery had a very curious
effect: once everything else was optimized, networking without GSO was
faster than networking with GSO, since more interrupts were sent and
hence a greater chance of one getting through to the Guest!

Note2: (Almost) Closing the same loophole for iret doesn't have any
measurable effect, so I'm leaving that patch for the moment.

Before:
	1GB tcpblast Guest->Host:		30.7 seconds
	1GB tcpblast Guest->Host (no GSO):	76.0 seconds

After:
	1GB tcpblast Guest->Host:		6.8 seconds
	1GB tcpblast Guest->Host (no GSO):	27.8 seconds

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
This commit is contained in:
Rusty Russell
2009-06-12 22:27:02 -06:00
parent abd41f037e
commit a32a8813d0
8 changed files with 43 additions and 16 deletions
Download Patch File Download Diff File Expand all files Collapse all files

21
arch/x86/lguest/boot.c
View File

@@ -205,6 +205,12 @@ PV_CALLEE_SAVE_REGS_THUNK(save_fl);
static void restore_fl(unsigned long flags)
{
lguest_data.irq_enabled = flags;
mb();
/* Null hcall forces interrupt delivery now, if irq_pending is
* set to X86_EFLAGS_IF (ie. an interrupt is pending, and flags
* enables interrupts. */
if (flags & lguest_data.irq_pending)
kvm_hypercall0(LHCALL_SEND_INTERRUPTS);
}
PV_CALLEE_SAVE_REGS_THUNK(restore_fl);
@@ -219,6 +225,11 @@ PV_CALLEE_SAVE_REGS_THUNK(irq_disable);
static void irq_enable(void)
{
lguest_data.irq_enabled = X86_EFLAGS_IF;
mb();
/* Null hcall forces interrupt delivery now. */
if (lguest_data.irq_pending)
kvm_hypercall0(LHCALL_SEND_INTERRUPTS);
}
PV_CALLEE_SAVE_REGS_THUNK(irq_enable);
@@ -972,10 +983,10 @@ static void lguest_restart(char *reason)
*
* Our current solution is to allow the paravirt back end to optionally patch
* over the indirect calls to replace them with something more efficient. We
* patch the four most commonly called functions: disable interrupts, enable
* interrupts, restore interrupts and save interrupts. We usually have 6 or 10
* bytes to patch into: the Guest versions of these operations are small enough
* that we can fit comfortably.
* patch two of the simplest of the most commonly called functions: disable
* interrupts and save interrupts. We usually have 6 or 10 bytes to patch
* into: the Guest versions of these operations are small enough that we can
* fit comfortably.
*
* First we need assembly templates of each of the patchable Guest operations,
* and these are in i386_head.S. */
@@ -986,8 +997,6 @@ static const struct lguest_insns
const char *start, *end;
} lguest_insns[] = {
[PARAVIRT_PATCH(pv_irq_ops.irq_disable)] = { lgstart_cli, lgend_cli },
[PARAVIRT_PATCH(pv_irq_ops.irq_enable)] = { lgstart_sti, lgend_sti },
[PARAVIRT_PATCH(pv_irq_ops.restore_fl)] = { lgstart_popf, lgend_popf },
[PARAVIRT_PATCH(pv_irq_ops.save_fl)] = { lgstart_pushf, lgend_pushf },
};