linux-kernel - [PATCH] x86, efi: Regression in Pass a minimal map to SetVirtualAddressMap()

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [day] [month] [year] [list]

Date:	Fri, 9 Sep 2011 09:18:21 -0500
From:	Russ Anderson <rja@....com>
To:	linux-kernel@...r.kernel.org, x86@...nel.org
Cc:	Matthew Garrett <mjg@...hat.com>,
	"H. Peter Anvin" <hpa@...ux.intel.com>,
	Russ Anderson <rja@....com>, stable@...nel.org
Subject: [PATCH] x86, efi: Regression in Pass a minimal map to SetVirtualAddressMap()

Commit 7cb00b72876ea2451eb79d468da0e8fb9134aa8a allocated a new memmap
table to be passed down to bios.  The problem is that there needs to be a 
physical mapping of the table for bios to access it.  The bottom 512Gig
of memory has a physical mapping, so if the table is in that range 
bios can access it.  If the table is above that range bios cannot.  
UV systems with more than 512Gig of memory have hit this problem.

This patch undoes the previous commit, reverting back to using the
old memmap table.

Reported-by: Russ Anderson <rja@....com>
Tested-by: Russ Anderson <rja@....com>
Signed-off-by: Russ Anderson <rja@....com>
Cc: stable@...nel.org


---
 arch/x86/platform/efi/efi.c |   14 +++-----------
 1 file changed, 3 insertions(+), 11 deletions(-)

Index: linux/arch/x86/platform/efi/efi.c
===================================================================
--- linux.orig/arch/x86/platform/efi/efi.c	2011-09-08 15:07:33.099462927 -0500
+++ linux/arch/x86/platform/efi/efi.c	2011-09-08 15:13:18.843970180 -0500
@@ -625,8 +625,7 @@ void __init efi_enter_virtual_mode(void)
 	efi_status_t status;
 	unsigned long size;
 	u64 end, systab, addr, npages, end_pfn;
-	void *p, *va, *new_memmap = NULL;
-	int count = 0;
+	void *p, *va;
 
 	efi.systab = NULL;
 
@@ -695,21 +694,15 @@ void __init efi_enter_virtual_mode(void)
 			systab += md->virt_addr - md->phys_addr;
 			efi.systab = (efi_system_table_t *) (unsigned long) systab;
 		}
-		new_memmap = krealloc(new_memmap,
-				      (count + 1) * memmap.desc_size,
-				      GFP_KERNEL);
-		memcpy(new_memmap + (count * memmap.desc_size), md,
-		       memmap.desc_size);
-		count++;
 	}
 
 	BUG_ON(!efi.systab);
 
 	status = phys_efi_set_virtual_address_map(
-		memmap.desc_size * count,
+		memmap.desc_size * memmap.nr_map,
 		memmap.desc_size,
 		memmap.desc_version,
-		(efi_memory_desc_t *)__pa(new_memmap));
+		memmap.phys_map);
 
 	if (status != EFI_SUCCESS) {
 		printk(KERN_ALERT "Unable to switch EFI into virtual mode "
@@ -747,7 +740,6 @@ void __init efi_enter_virtual_mode(void)
 		runtime_code_page_mkexec();
 	early_iounmap(memmap.map, memmap.nr_map * memmap.desc_size);
 	memmap.map = NULL;
-	kfree(new_memmap);
 }
 
 /*
-- 
Russ Anderson, OS RAS/Partitioning Project Lead  
SGI - Silicon Graphics Inc          rja@....com
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/