diff options
| author | cinap_lenrek <cinap_lenrek@felloff.net> | 2019-09-14 14:02:34 +0200 |
|---|---|---|
| committer | cinap_lenrek <cinap_lenrek@felloff.net> | 2019-09-14 14:02:34 +0200 |
| commit | acab8881bc2c0befcda62c2e6cc479cd7529dbe3 (patch) | |
| tree | 8e5013e2fd5890fd63544f422b8f05fe0bbfd0f3 /sys/src/9/bcm64/mmu.c | |
| parent | ca2f1c07f2a6bb48582b990868976b45d3663605 (diff) | |
bcm64: enter page tables in mmutop *AFTER* switching asid in mmuswitch()
there was a small window between modifying mmutop and switching the
asid where the core could bring in the new entries under the old asid
into the tlb due to speculation / prefetching.
this change moves the entering of the page tables into mmutop after
setttbr() to prevent this scenario.
due to us switching to the resereved asid 0 on procsave()->putasid(),
the only asid that could have potentially been poisoned would be asid 0
which does not have any user mappings. so this did not show any noticable
effect.
Diffstat (limited to 'sys/src/9/bcm64/mmu.c')
| -rw-r--r-- | sys/src/9/bcm64/mmu.c | 10 |
1 files changed, 5 insertions, 5 deletions
diff --git a/sys/src/9/bcm64/mmu.c b/sys/src/9/bcm64/mmu.c index ca8186322..e53270af9 100644 --- a/sys/src/9/bcm64/mmu.c +++ b/sys/src/9/bcm64/mmu.c @@ -499,15 +499,15 @@ mmuswitch(Proc *p) p->newtlb = 0; } - for(t = p->mmuhead[PTLEVELS-1]; t != nil; t = t->next){ - va = t->va; - m->mmutop[PTLX(va, PTLEVELS-1)] = t->pa | PTEVALID | PTETABLE; - } - if(allocasid(p)) flushasid((uvlong)p->asid<<48); setttbr((uvlong)p->asid<<48 | PADDR(m->mmutop)); + + for(t = p->mmuhead[PTLEVELS-1]; t != nil; t = t->next){ + va = t->va; + m->mmutop[PTLX(va, PTLEVELS-1)] = t->pa | PTEVALID | PTETABLE; + } } void |