mm: fix a race condition under SMC + COW
Subsystems (2)
PATCH 1/1 mm: fix a race condition under SMC + COW
From: "Siddha, Suresh B" <[email protected]>
Failing context is a multi threaded process context and the failing sequence is as follows.
One thread T0 doing self modifying code on page X on processor P0 and another thread T1 doing COW (breaking the COW setup as part of just happened fork() in another thread T2) on the same page X on processor P1. T0 doing SMC can endup modifying the new page Y (allocated by the T1 doing COW on P1) but because of different I/D TLB's, P0 ITLB will not see the new mapping till the flush TLB IPI from P1 is received. During this interval, if T0 executes the code created by SMC it can result in an app error (as ITLB still points to old page X and endup executing the content in page X rather than using the content in page Y).
Fix this issue by first clearing the PTE and flushing it, before updating it with new entry.
---
mm/memory.c renamed from mm/memory.c.orig +8 -1
@@ -1551,7 +1551,14 @@ gotten: entry =mk_pte (new_page, vma->vm_page_prot); entry =maybe_mkwrite (pte_mkdirty (entry), vma);lazy_mmu_prot_update (entry); -ptep_establish (vma, address, page_table, entry); +/* +* Clear the pte entry and flush it first, before updating the +* pte with the new entry. This will avoid a race condition +* seen in the presence of one thread doing SMC and another +* thread doing COW. +*/ +ptep_clear_flush (vma, address, page_table); +set_pte_at (mm, address, page_table, entry);update_mmu_cache (vma, address, entry);lru_cache_add_active (new_page);page_add_new_anon_rmap (new_page, vma, address);
From: "Siddha, Suresh B" <[email protected]>
Date: Wed, 27 Sep 2006 15:15:07 -0700
Quoted 15 lines
From: "Siddha, Suresh B" <[email protected]>
Failing context is a multi threaded process context and the failing sequence is as follows.
One thread T0 doing self modifying code on page X on processor P0 and another thread T1 doing COW (breaking the COW setup as part of just happened
fork()in another thread T2) on the same page X on processor P1. T0 doing SMC can endup modifying the new page Y (allocated by the T1 doing COW on P1) but because of different I/D TLB's, P0 ITLB will not see the new mapping till the flush TLB IPI from P1 is received. During this interval, if T0 executes the code created by SMC it can result in an app error (as ITLB still points to old page X and endup executing the content in page X rather than using the content in page Y).
Quoted 4 lines
Fix this issue by first clearing the PTE and flushing it, before updating it with new entry.
Signed-off-by Suresh Siddha <[email protected]>
You can't really do a set_pte_at() in this code path because there isn't a subsequent flush_tlb_*().
This is needed because some architectures queue up all set_pte_at() calls until the next flush_tlb_*() in order to batch TLB flushes. PowerPC and Sparc64 both do this.
The pte_establish() in the existing code works fine because it takes care of the set_pte_at() and flush_tlb_*() work internally when necessary on a given platform.
Quoted 10 lines · David Miller
You can't really do a
set_pte_at()in this code path because there isn't a subsequent flush_tlb_*().This is needed because some architectures queue up all
set_pte_at()calls until the next flush_tlb_*() in order to batch TLB flushes. PowerPC and Sparc64 both do this.The
pte_establish()in the existing code works fine because it takes care of theset_pte_at()and flush_tlb_*() work internally when necessary on a given platform.
I am flushing the pte entry in ptep_clear_flush() and it is Ok not to do another TLB flush after doing set_pte_at().
On Sparc64, this new set_pte_at() (after ptep_clear_flush) will not batch any TLB flush as the previous pte contents were zero.
We are Ok with this patch, isn't it?
thanks,
suresh
From: "Siddha, Suresh B" <[email protected]>
Date: Wed, 27 Sep 2006 17:23:55 -0700
Quoted 7 lines
I am flushing the pte entry in
ptep_clear_flush()and it is Ok not to do another TLB flush after doingset_pte_at().On Sparc64, this new
set_pte_at()(afterptep_clear_flush) will not batch any TLB flush as the previous pte contents were zero.We are Ok with this patch, isn't it?
Ok, it seems PowerPC also has the "don't do anything if previous PTE was zero" logic. So yes, it should be ok.
Quoted 15 lines · Suresh B
From: "Siddha, Suresh B" <[email protected]>
Failing context is a multi threaded process context and the failing sequence is as follows.
One thread T0 doing self modifying code on page X on processor P0 and another thread T1 doing COW (breaking the COW setup as part of just happened
fork()in another thread T2) on the same page X on processor P1. T0 doing SMC can endup modifying the new page Y (allocated by the T1 doing COW on P1) but because of different I/D TLB's, P0 ITLB will not see the new mapping till the flush TLB IPI from P1 is received. During this interval, if T0 executes the code created by SMC it can result in an app error (as ITLB still points to old page X and endup executing the content in page X rather than using the content in page Y).
Quoted 4 lines
Fix this issue by first clearing the PTE and flushing it, before updating it with new entry.
Signed-off-by Suresh Siddha <[email protected]>
I was a bit sceptical, in the habit of thinking that Self Modifying Code must look such issues itself: but I guess there's nothing it can do to avoid this one.
Fair enough, what you're changing it to is pretty much what powerpc and s390 were already doing, and is a more robust way of proceeding, consistent with how ptes are set everywhere else.
The ptep_clear_flush is a bit heavy-handed (it's anxious to return the pte that was atomically cleared), but we'd have to wander through lots of arches to get the right minimal behaviour. It'd also be nice to eliminate ptep_establish completely, now only used to define other macros/inlines: it always seemed obfuscation to me, what you've got there now is clearer. Let's put those cleanups on a TODO list.
Quoted 20 lines
---
mm/memory.c renamed from mm/memory.c.orig +8
-1@@ -1551,7 +1551,14 @@ gotten: entry =mk_pte (new_page, vma->vm_page_prot); entry =maybe_mkwrite (pte_mkdirty (entry), vma);lazy_mmu_prot_update (entry); -ptep_establish (vma, address, page_table, entry); +/* +* Clear the pte entry and flush it first, before updating the +* pte with the new entry. This will avoid a race condition +* seen in the presence of one thread doing SMC and another +* thread doing COW. +*/ +ptep_clear_flush (vma, address, page_table); +set_pte_at (mm, address, page_table, entry);update_mmu_cache (vma, address, entry);lru_cache_add_active (new_page);page_add_new_anon_rmap (new_page, vma, address);