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>
When the Guest does the LHCALL_HALT hypercall, we go to sleep, expecting
that a timer or the Waker will wake_up_process() us.
But we do it in a stupid way, leaving a classic missing wakeup race.
So split maybe_do_interrupt() into interrupt_pending() and
try_deliver_interrupt(), and check maybe_do_interrupt() and the
"break_out" flag before calling schedule.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
The Launcher could be inside the Guest on another CPU; wake_up_process
will do nothing because it is "running". kick_process will knock it
back into our kernel in this case, otherwise we'll miss it until the
next guest exit.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Impact: clean up
Rusty told me, some time ago, that he had become a fan of "bool".
So, here are some replacements.
Signed-off-by: Matias Zabaljauregui <zabaljauregui at gmail.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Impact: cleanup
This patch allow us to use KVM hypercalls
Signed-off-by: Matias Zabaljauregui <zabaljauregui at gmail.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Impact: fix lguest, clean up
32-bit lguest used used_vectors to record vectors, but that model of
allocating vectors changed and got broken, after we changed vector
allocation to a per_cpu array.
Try enable that for 64bit, and the array is used for all vectors that
are not managed by vector_irq per_cpu array.
Also kill system_vectors[], that is now a duplication of the
used_vectors bitmap.
[ merged in cpus4096 due to io_apic.c cpumask changes. ]
[ -v2, fix build failure ]
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Ron Minnich noticed that guest userspace gets a GPF when it tries to int3:
we need to copy the privilege level from the guest-supplied IDT to the real
IDT. int3 is the only common case where guest userspace expects to invoke
an interrupt, so that's the symptom of failing to do this.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Took some cycles to re-read the Lguest Journey end-to-end, fix some
rot and tighten some phrases.
Only comments change. No new jokes, but a couple of recycled old jokes.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
We can save some lines of code by getting rid of
*lg = cpu... lines of code spread everywhere by now.
Signed-off-by: Glauber de Oliveira Costa <gcosta@redhat.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
events represented in the 'changed' bitmap are per-cpu, not per-guest.
move it to the lg_cpu structure
Signed-off-by: Glauber de Oliveira Costa <gcosta@redhat.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
this patch makes the pgdir management per-vcpu. The pgdirs pool
is still guest-wide (although it'll probably need to grow when we
are really executing more vcpus), but the pgdidx index is gone,
since it makes no sense anymore. Instead, we use a per-vcpu
index.
Signed-off-by: Glauber de Oliveira Costa <gcosta@redhat.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
lguest struct have room for some fields, namely, cr2, ts, esp1
and ss1, that are not really guest-wide, but rather, vcpu-wide.
This patch puts it in the vcpu struct
Signed-off-by: Glauber de Oliveira Costa <gcosta@redhat.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
lguest uses tasks to control its running behaviour (like sending
breaks, controlling halted state, etc). In a per-vcpu environment,
each vcpu will have its own underlying task. So this patch
makes the infrastructure for that possible
Signed-off-by: Glauber de Oliveira Costa <gcosta@redhat.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
The fields found in lguest_arch are not really per-guest,
but per-cpu (gdt, idt, etc). So this patch turns lguest_arch
into lg_cpu_arch.
It makes sense to have a per-guest per-arch struct, but this
can be addressed later, when the need arrives.
Signed-off-by: Glauber de Oliveira Costa <gcosta@redhat.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
This is the most obvious per-vcpu field: registers.
So this patch moves it from struct lguest to struct vcpu,
and patch the places in which they are used, accordingly
Signed-off-by: Glauber de Oliveira Costa <gcosta@redhat.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
This patch adapts interrupt processing for using the vcpu struct.
Signed-off-by: Glauber de Oliveira Costa <gcosta@redhat.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Here, I introduce per-vcpu timers. With this, we can have
local expiries, needed for accounting time in smp guests
Signed-off-by: Glauber de Oliveira Costa <gcosta@redhat.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Went through the documentation doing typo and content fixes. This
patch contains only comment and whitespace changes.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Jes complains that page table code still uses lgread_u32 even though
it now uses general kernel pte types. The best thing to do is to
generalize lgread_u32 and lgwrite_u32.
This means we lose the efficiency of getuser(). We could potentially
regain it if we used __copy_from_user instead of copy_from_user, but
I'm not certain that our range check is equivalent to access_ok() on
all platforms.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Acked-by: Jes Sorensen <jes@sgi.com>
1) This allows us to get alot closer to booting bzImages.
2) It means we don't have to know page_offset.
3) The Guest needs to modify the boot pagetables to create the
PAGE_OFFSET mapping before jumping to C code.
4) guest_pa() walks the page tables rather than using page_offset.
5) We don't use page_offset to figure out whether to emulate: it was
always kinda quesationable, and won't work for instructions done
before remapping (bzImage unpacking in particular).
6) We still want the kernel address for tlb flushing: have the initial
hypercall give us that, too.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
(Based on Ron Minnich's LGUEST_PLAN9_SYSCALL patch).
This patch allows Guests to specify what system call vector they want,
and we try to reserve it. We only allow one non-Linux system call
vector, to try to avoid DoS on the Host.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Separate i386 architecture specific from core.c and move it to
x86/core.c and add x86/lguest.h header file to match.
Signed-off-by: Jes Sorensen <jes@sgi.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
This simplifies the code a little, in preparation for allowing
alternate system call vectors in guests (Plan 9 uses 0x40).
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
If the stack pointer is 0xc057a000, then the first stack page is at
0xc0579000 (the stack pointer is decremented before use). Not
calculating this correctly caused guests with CONFIG_DEBUG_PAGEALLOC=y
to be killed with a "bad stack page" message: the initial kernel stack
was just proceeding the .smp_locks section which
CONFIG_DEBUG_PAGEALLOC marks read-only when freeing.
Thanks to Frederik Deweerdt for the bug report!
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
If a Guest makes hypercall which sets a GDT entry to not present, we
currently set any segment registers using that GDT entry to 0.
Unfortunately, this is not sufficient: there are other ways of
altering GDT entries which will cause a fault.
The correct solution to do what Linux does: let them set any GDT value
they want and handle the #GP when popping causes a fault. This has
the added benefit of making our Switcher slightly more robust in the
case of any other bugs which cause it to fault.
We kill the Guest if it causes a fault in the Switcher: it's the
Guest's responsibility to make sure it's not using segments when it
changes them.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
A non-periodic clock_event_device and the "jiffies" clock don't mix well:
tick_handle_periodic() can go into an infinite loop.
Currently lguest guests use the jiffies clock when the TSC is
unusable. Instead, make the Host write the current time into the lguest
page on every interrupt. This doesn't cost much but is more precise
and at least as accurate as the jiffies clock. It also gets rid of
the GET_WALLCLOCK hypercall.
Also, delay setting sched_clock until our clock is set up, otherwise
the early printk timestamps can go backwards (not harmful, just ugly).
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Documentation: The FIXMEs
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Documentation: The Host
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The netfilter code had very good documentation: the Netfilter Hacking HOWTO.
Noone ever read it.
So this time I'm trying something different, using a bit of Knuthiness.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The sense of the IF bit is backwards in the host interrupt handling.
This means we always save "IF=1" on the stack when injecting an
interrupt. It turns out this is almost always correct (unless the
guest is taking a page fault in an interrupt due to an unpopulated
vmalloc mapping), so went unnoticed.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This is the code for the "lg.ko" module, which allows lguest guests to
be launched.
[akpm@linux-foundation.org: update for futex-new-private-futexes]
[akpm@linux-foundation.org: build fix]
[jmorris@namei.org: lguest: use hrtimers]
[akpm@linux-foundation.org: x86_64 build fix]
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Cc: Andi Kleen <ak@suse.de>
Cc: Eric Dumazet <dada1@cosmosbay.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>