Xen devel - Feb 2012 - [PATCH] Disable PAT support when running under Xen (v1).

The issue at hand is that any prolonged usage of radeon or nouveau driver
ends up corrupting the file system or we end up with mysterious crashes of
applications.

There are three ways of fixing it:
 a). A proper fix: https://lkml.org/lkml/2012/2/10/228 . I posted the same
   fix for 3.2 way back in December but it got nowhere. The recent posting
   has also been meet with silence. Not being happy with this laying
   around and corrupting folks file systems (and in my case my git tree):

 b). Disable PAT when run under Xen. At least until we get the a) moving
   or alternative solutions are found.

 c). Use the ''nopat'' argument on the Linux command line (which
is what
   this patch set (b) forces unconditionally) and has been recommended on the
   BZs for users.

In summary this patch set:
  - fixes corruption issues
  - makes graphic drivers not able to use Write-Combine - meaning you
can''t
    get super-fast performance using graphic drivers.

Konrad Rzeszutek Wilk (2):
      xen/setup: Remove redundant filtering of PTE masks.
      xen/pat: Disable PAT support for now.

 arch/x86/xen/enlighten.c |    6 ++----
 arch/x86/xen/mmu.c       |    8 ++++----
 2 files changed, 6 insertions(+), 8 deletions(-)


diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c
index 12eb07b..4172af8 100644
--- a/arch/x86/xen/enlighten.c
+++ b/arch/x86/xen/enlighten.c
@@ -1141,7 +1141,9 @@ asmlinkage void __init xen_start_kernel(void)
 
 	/* Prevent unwanted bits from being set in PTEs. */
 	__supported_pte_mask &= ~_PAGE_GLOBAL;
+#if 0
 	if (!xen_initial_domain())
+#endif
 		__supported_pte_mask &= ~(_PAGE_PWT | _PAGE_PCD);
 
 	__supported_pte_mask |= _PAGE_IOMAP;
@@ -1204,10 +1206,6 @@ asmlinkage void __init xen_start_kernel(void)
 
 	pgd = (pgd_t *)xen_start_info->pt_base;
 
-	if (!xen_initial_domain())
-		__supported_pte_mask &= ~(_PAGE_PWT | _PAGE_PCD);
-
-	__supported_pte_mask |= _PAGE_IOMAP;
 	/* Don''t do the full vcpu_info placement stuff until we have a
 	   possible map and a non-dummy shared_info. */
 	per_cpu(xen_vcpu, 0) = &HYPERVISOR_shared_info->vcpu_info[0];
diff --git a/arch/x86/xen/mmu.c b/arch/x86/xen/mmu.c
index 58a0e46..95c1cf6 100644
--- a/arch/x86/xen/mmu.c
+++ b/arch/x86/xen/mmu.c
@@ -415,13 +415,13 @@ static pteval_t iomap_pte(pteval_t val)
 static pteval_t xen_pte_val(pte_t pte)
 {
 	pteval_t pteval = pte.pte;
-
+#if 0
 	/* If this is a WC pte, convert back from Xen WC to Linux WC */
 	if ((pteval & (_PAGE_PAT | _PAGE_PCD | _PAGE_PWT)) == _PAGE_PAT) {
 		WARN_ON(!pat_enabled);
 		pteval = (pteval & ~_PAGE_PAT) | _PAGE_PWT;
 	}
-
+#endif
 	if (xen_initial_domain() && (pteval & _PAGE_IOMAP))
 		return pteval;
 
@@ -463,7 +463,7 @@ void xen_set_pat(u64 pat)
 static pte_t xen_make_pte(pteval_t pte)
 {
 	phys_addr_t addr = (pte & PTE_PFN_MASK);
-
+#if 0
 	/* If Linux is trying to set a WC pte, then map to the Xen WC.
 	 * If _PAGE_PAT is set, then it probably means it is really
 	 * _PAGE_PSE, so avoid fiddling with the PAT mapping and hope
@@ -476,7 +476,7 @@ static pte_t xen_make_pte(pteval_t pte)
 		if ((pte & (_PAGE_PCD | _PAGE_PWT)) == _PAGE_PWT)
 			pte = (pte & ~(_PAGE_PCD | _PAGE_PWT)) | _PAGE_PAT;
 	}
-
+#endif
 	/*
 	 * Unprivileged domains are allowed to do IOMAPpings for
 	 * PCI passthrough, but not map ISA space.  The ISA

commit 7347b4082e55ac4a673f06a0a0ce25c37273c9ec "xen: Allow
unprivileged Xen domains to create iomap pages" added a redundant
line in the early bootup code to filter out the PTE. That
filtering is already done a bit earlier so this extra processing
is not required.

Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
---
 arch/x86/xen/enlighten.c |    4 ----
 1 files changed, 0 insertions(+), 4 deletions(-)

diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c
index 12eb07b..7c44e1b 100644
--- a/arch/x86/xen/enlighten.c
+++ b/arch/x86/xen/enlighten.c
@@ -1204,10 +1204,6 @@ asmlinkage void __init xen_start_kernel(void)
 
 	pgd = (pgd_t *)xen_start_info->pt_base;
 
-	if (!xen_initial_domain())
-		__supported_pte_mask &= ~(_PAGE_PWT | _PAGE_PCD);
-
-	__supported_pte_mask |= _PAGE_IOMAP;
 	/* Don''t do the full vcpu_info placement stuff until we have a
 	   possible map and a non-dummy shared_info. */
 	per_cpu(xen_vcpu, 0) = &HYPERVISOR_shared_info->vcpu_info[0];
-- 
1.7.9.48.g85da4d

[Pls also look at https://lkml.org/lkml/2012/2/10/228]

Using of PAT to change pages from WB to WC works quite nicely.
Changing it back to WB - not so much. The crux of the matter is
that the code that does this (__page_change_att_set_clr) has only
limited information so when it tries to the change it gets
the "raw" unfiltered information instead of the properly filtered one
-
and the "raw" one tell it that PSE bit is on (while infact it
is not).  As a result when the PTE is set to be WB from WC, we get
tons of:

:WARNING: at arch/x86/xen/mmu.c:475 xen_make_pte+0x67/0xa0()
:Hardware name: HP xw4400 Workstation
.. snip..
:Pid: 27, comm: kswapd0 Tainted: G        W    3.2.2-1.fc16.x86_64 #1
:Call Trace:
: [<ffffffff8106dd1f>] warn_slowpath_common+0x7f/0xc0
: [<ffffffff8106dd7a>] warn_slowpath_null+0x1a/0x20
: [<ffffffff81005a17>] xen_make_pte+0x67/0xa0
: [<ffffffff810051bd>] __raw_callee_save_xen_make_pte+0x11/0x1e
: [<ffffffff81040e15>] ? __change_page_attr_set_clr+0x9d5/0xc00
: [<ffffffff8114c2e8>] ? __purge_vmap_area_lazy+0x158/0x1d0
: [<ffffffff8114cca5>] ? vm_unmap_aliases+0x175/0x190
: [<ffffffff81041168>] change_page_attr_set_clr+0x128/0x4c0
: [<ffffffff81041542>] set_pages_array_wb+0x42/0xa0
: [<ffffffff8100a9b2>] ? check_events+0x12/0x20
: [<ffffffffa0074d4c>] ttm_pages_put+0x1c/0x70 [ttm]
: [<ffffffffa0074e98>] ttm_page_pool_free+0xf8/0x180 [ttm]
: [<ffffffffa0074f78>] ttm_pool_mm_shrink+0x58/0x90 [ttm]
: [<ffffffff8112ba04>] shrink_slab+0x154/0x310
: [<ffffffff8112f17a>] balance_pgdat+0x4fa/0x6c0
: [<ffffffff8112f4b8>] kswapd+0x178/0x3d0
: [<ffffffff815df134>] ? __schedule+0x3d4/0x8c0
: [<ffffffff81090410>] ? remove_wait_queue+0x50/0x50
: [<ffffffff8112f340>] ? balance_pgdat+0x6c0/0x6c0
: [<ffffffff8108fb6c>] kthread+0x8c/0xa0

for every page. The fix for this is has been posted
and is https://lkml.org/lkml/2012/2/10/228
"x86/cpa: Use pte_attrs instead of pte_flags on CPA/set_p.._wb/wc
operations."
along with a detailed description of the problem and solution.

But since that posting has gone nowhere I am proposing
this band-aid solution so that at least users don''t get
the page corruption (the pages that are WC don''t get changed to WB
and end up being recycled for filesystem or other things causing
mysterious crashes).

The negative impact of this patch is that users of WC flag
(which are InfiniBand, radeon, nouveau drivers) won''t be able
to set that flag - so they are going to see performance degradation.
But stability is more important here.

Fixes RH BZ# 742032, 787403, and 745574
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
---
 arch/x86/xen/enlighten.c |    2 ++
 arch/x86/xen/mmu.c       |    8 ++++----
 2 files changed, 6 insertions(+), 4 deletions(-)

diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c
index 7c44e1b..4172af8 100644
--- a/arch/x86/xen/enlighten.c
+++ b/arch/x86/xen/enlighten.c
@@ -1141,7 +1141,9 @@ asmlinkage void __init xen_start_kernel(void)
 
 	/* Prevent unwanted bits from being set in PTEs. */
 	__supported_pte_mask &= ~_PAGE_GLOBAL;
+#if 0
 	if (!xen_initial_domain())
+#endif
 		__supported_pte_mask &= ~(_PAGE_PWT | _PAGE_PCD);
 
 	__supported_pte_mask |= _PAGE_IOMAP;
diff --git a/arch/x86/xen/mmu.c b/arch/x86/xen/mmu.c
index 58a0e46..95c1cf6 100644
--- a/arch/x86/xen/mmu.c
+++ b/arch/x86/xen/mmu.c
@@ -415,13 +415,13 @@ static pteval_t iomap_pte(pteval_t val)
 static pteval_t xen_pte_val(pte_t pte)
 {
 	pteval_t pteval = pte.pte;
-
+#if 0
 	/* If this is a WC pte, convert back from Xen WC to Linux WC */
 	if ((pteval & (_PAGE_PAT | _PAGE_PCD | _PAGE_PWT)) == _PAGE_PAT) {
 		WARN_ON(!pat_enabled);
 		pteval = (pteval & ~_PAGE_PAT) | _PAGE_PWT;
 	}
-
+#endif
 	if (xen_initial_domain() && (pteval & _PAGE_IOMAP))
 		return pteval;
 
@@ -463,7 +463,7 @@ void xen_set_pat(u64 pat)
 static pte_t xen_make_pte(pteval_t pte)
 {
 	phys_addr_t addr = (pte & PTE_PFN_MASK);
-
+#if 0
 	/* If Linux is trying to set a WC pte, then map to the Xen WC.
 	 * If _PAGE_PAT is set, then it probably means it is really
 	 * _PAGE_PSE, so avoid fiddling with the PAT mapping and hope
@@ -476,7 +476,7 @@ static pte_t xen_make_pte(pteval_t pte)
 		if ((pte & (_PAGE_PCD | _PAGE_PWT)) == _PAGE_PWT)
 			pte = (pte & ~(_PAGE_PCD | _PAGE_PWT)) | _PAGE_PAT;
 	}
-
+#endif
 	/*
 	 * Unprivileged domains are allowed to do IOMAPpings for
 	 * PCI passthrough, but not map ISA space.  The ISA
-- 
1.7.9.48.g85da4d