linux/arch/s390/kvm
Christian Borntraeger 1365039d0c KVM: s390: Fix ipte locking
ipte_unlock_siif uses cmpxchg to replace the in-memory data of the ipte
lock together with ACCESS_ONCE for the intial read.

union ipte_control {
        unsigned long val;
        struct {
                unsigned long k  : 1;
                unsigned long kh : 31;
                unsigned long kg : 32;
        };
};
[...]
static void ipte_unlock_siif(struct kvm_vcpu *vcpu)
{
        union ipte_control old, new, *ic;

        ic = &vcpu->kvm->arch.sca->ipte_control;
        do {
                new = old = ACCESS_ONCE(*ic);
                new.kh--;
                if (!new.kh)
                        new.k = 0;
        } while (cmpxchg(&ic->val, old.val, new.val) != old.val);
        if (!new.kh)
                wake_up(&vcpu->kvm->arch.ipte_wq);
}

The new value, is loaded twice from memory with gcc 4.7.2 of
fedora 18, despite the ACCESS_ONCE:

--->

l       %r4,0(%r3)      <--- load first 32 bit of lock (k and kh) in r4
alfi    %r4,2147483647  <--- add -1 to r4
llgtr   %r4,%r4         <--- zero out the sign bit of r4
lg      %r1,0(%r3)      <--- load all 64 bit of lock into new
lgr     %r2,%r1         <--- load the same into old
risbg   %r1,%r4,1,31,32 <--- shift and insert r4 into the bits 1-31 of
new
llihf   %r4,2147483647
ngrk    %r4,%r1,%r4
jne     aa0 <ipte_unlock+0xf8>
nihh    %r1,32767
lgr     %r4,%r2
csg     %r4,%r1,0(%r3)
cgr     %r2,%r4
jne     a70 <ipte_unlock+0xc8>

If the memory value changes between the first load (l) and the second
load (lg) we are broken. If that happens VCPU threads will hang
(unkillable) in handle_ipte_interlock.

Andreas Krebbel analyzed this and tracked it down to a compiler bug in
that version:
"while it is not that obvious the C99 standard basically forbids
duplicating the memory access also in that case. For an argumentation of
a similiar case please see:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=22278#c43

For the implementation-defined cases regarding volatile there are some
GCC-specific clarifications which can be found here:
https://gcc.gnu.org/onlinedocs/gcc/Volatiles.html#Volatiles

I've tracked down the problem with a reduced testcase. The problem was
that during a tree level optimization (SRA - scalar replacement of
aggregates) the volatile marker is lost. And an RTL level optimizer (CSE
- common subexpression elimination) then propagated the memory read into
  its second use introducing another access to the memory location. So
indeed Christian's suspicion that the union access has something to do
with it is correct (since it triggered the SRA optimization).

This issue has been reported and fixed in the GCC 4.8 development cycle:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=58145"

This patch replaces the ACCESS_ONCE scheme with a barrier() based scheme
that should work for all supported compilers.

Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Cc: stable@vger.kernel.org # v3.16+
2014-11-07 11:10:26 +01:00
..
diag.c KVM: s390/cmm: Fix prefix handling for diag 10 balloon 2014-09-10 12:19:42 +02:00
gaccess.c KVM: s390: Fix ipte locking 2014-11-07 11:10:26 +01:00
gaccess.h KVM: s390: Enable DAT support for TPROT handler 2014-05-30 09:39:36 +02:00
guestdbg.c KVM: s390: return -EFAULT if copy_from_user() fails 2014-05-06 14:57:59 +02:00
intercept.c KVM: s390: implement KVM_(S|G)ET_MP_STATE for user space state control 2014-07-10 14:11:17 +02:00
interrupt.c KVM: s390: count vcpu wakeups in stat.halt_wakeup 2014-10-01 14:42:14 +02:00
irq.h KVM: s390: irq routing for adapter interrupts. 2014-03-21 13:43:00 +01:00
Kconfig KVM: Give IRQFD its own separate enabling Kconfig option 2014-08-05 14:26:28 +02:00
kvm-s390.c KVM: s390: count vcpu wakeups in stat.halt_wakeup 2014-10-01 14:42:14 +02:00
kvm-s390.h KVM: s390: register flic ops dynamically 2014-09-17 13:10:09 +02:00
Makefile KVM: s390: hardware support for guest debugging 2014-04-22 13:24:51 +02:00
priv.c KVM: s390/mm: cleanup gmap function arguments, variable names 2014-08-25 14:35:58 +02:00
sigp.c KVM: s390: Fix memory leak on busy SIGP stop 2014-07-30 15:29:40 +02:00
trace-s390.h KVM: s390: enable IBS for single running VCPUs 2014-04-29 15:01:54 +02:00
trace.h KVM: s390: interpretive execution of SIGP EXTERNAL CALL 2014-05-16 14:57:28 +02:00