Skip to contentSkip to conversation

: Clash with ioremap_nocache() + _PAGE_PWT

Lguest/x86 In review Ahmed S. Darwish 3 participants 5 messages Feb 6, 2008 – Mar 4, 2008

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

Hi all,

Beginning from commit 4138cc3418f5, ioremap_nocache() sets the _PAGE_PWT flag.

Lguest doesn't accept a guest pte with a _PWT flag and reports a "bad page table entry" in that case.

I've removed check from lguest code and everything worked fine. Is this safe from the Lguest side ?

Mentioned commit [*]:

commit 4138cc3418f5eaa7524ff8e927102863f1ba0ea5

Author: Siddha, Suresh B <[email protected]>
Date:   Wed Jan 30 13:33:43 2008 +0100

x86: set strong uncacheable where UC is really desired

Also use _PAGE_PWT for all the mappings which need uncache mapping. Instead of existing PAT2 which is UC- (and can be overwritten by MTRRs), we now use PAT3 which is strong uncacheable.

This makes it consistent with pgprot_noncached()

arch/x86/mm/ioremap_32.c +1 -1
index 0b27831..ef0f6a4 100644
@@ -119,7 +119,7 @@ EXPORT_SYMBOL(__ioremap);
 void __iomem *ioremap_nocache (unsigned long phys_addr, unsigned long size)
 {
 	unsigned long last_addr;
-	void __iomem *p = __ioremap(phys_addr, size, _PAGE_PCD);
+	void __iomem *p = __ioremap(phys_addr, size, _PAGE_PCD | _PAGE_PWT);
 	if (!p) 
 		return p; 
 

Thanks,

[*]: latest pull calls set_memory_uc() which also sets the _PWT flag.

--
Ahmed S. Darwish
Homepage: http://darwish.07.googlepages.com
Blog: http://darwish-07.blogspot.com

#1 lore raw
Quoted 10 lines · Ahmed S. Darwish

Hi all,

Beginning from commit 4138cc3418f5, ioremap_nocache() sets the _PAGE_PWT flag.

Lguest doesn't accept a guest pte with a _PWT flag and reports a "bad page table entry" in that case.

I've removed check from lguest code and everything worked fine. Is this safe from the Lguest side ?

yes, should be safe. Could you send a patch so that others can apply it too?

Ingo

#2 lore raw

PATCH 1/1 lguest: Accept guest _PAGE_PWT page table entries

Quoted 8 lines · Ingo Molnar
Quoted 5 lines · Ahmed S. Darwish

Hi all,

Beginning from commit 4138cc3418f5, ioremap_nocache() sets the _PAGE_PWT flag.

Quoted 9 lines
Quoted 5 lines

Lguest doesn't accept a guest pte with a _PWT flag and reports a "bad page table entry" in that case.

I've removed check from lguest code and everything worked fine. Is this safe from the Lguest side ?

yes, should be safe. Could you send a patch so that others can apply it too?

Ofcourse :) :

Accept guest _PAGE_PWT page table entries, otherwise lguest will always fail with a "bad page table entry" message.

---

drivers/lguest/page_tables.c +2 -2
index 74b4cf2..952160b 100644
@@ -178,8 +178,8 @@ static void release_pte(pte_t pte)
 
 static void check_gpte(struct lg_cpu *cpu, pte_t gpte)
 {
-	if ((pte_flags(gpte) & (_PAGE_PWT|_PAGE_PSE))
-	    || pte_pfn(gpte) >= cpu->lg->pfn_limit)
+	if ((pte_flags(gpte) & _PAGE_PSE) || 
+	    pte_pfn(gpte) >= cpu->lg->pfn_limit)
 		kill_guest(cpu, "bad page table entry");
 }
 

Regards,

--
Ahmed S. Darwish
Homepage: http://darwish.07.googlepages.com
Blog: http://darwish-07.blogspot.com

#3 lore raw

Re: [PATCH] lguest: Accept guest _PAGE_PWT page table entries

Quoted 4 lines · Ahmed S. Darwish

Ofcourse :) :

Accept guest _PAGE_PWT page table entries, otherwise lguest will always fail with a "bad page table entry" message.

thanks - i've queued this up into x86.git. Rusty, is this fine to you, and should we nurse this fix upstream?

Ingo

#4 lore raw

Re: [PATCH] lguest: Accept guest _PAGE_PWT page table entries

In reply to Ahmed S. Darwish
Quoted 10 lines · On Thursday 07 February 2008 11:51:20 Ahmed S. Darwish wrote:
Quoted 2 lines · Ingo Molnar

yes, should be safe. Could you send a patch so that others can apply it too?

Ofcourse :) :

Accept guest _PAGE_PWT page table entries, otherwise lguest will always fail with a "bad page table entry" message.

Thanks for tracking this Ahmed, and thanks to Ingo for pushing it upstream.

Cheers,
Rusty.