Skip to contentSkip to conversation

tidy AMD 768MPX fix

PATCH In review Andy Whitcroft 3 participants 4 messages Sep 3, 2004 – Sep 6, 2004

lore
Share & tools
lore

Andy Whitcroft

author
#0 lore raw

PATCH 1/1 tidy AMD 768MPX fix

Convert the AMD 768MPX errata #56 fix to use PAGE_SIZE instead of using 4096 in line with other declarations in this file. Also take the oppotunity to match indentation.

Revision: $Rev: 612 $

Signed-off-by Andy Whitcroft <[email protected]>

diffstat 020-tidy-AMD-768MPX-fix

---
 setup.c |   10 ++++++----
 1 files changed, 6 insertions(+), 4 deletions(-)
arch/i386/kernel/setup.c +6 -4
diff -X /home/apw/brief/lib/vdiff.excl -rupN reference/arch/i386/kernel/setup.c current/arch/i386/kernel/setup.c
@@ -1052,12 +1052,14 @@ static unsigned long __init setup_memory
 	/* reserve EBDA region, it's a 4K region */
 	reserve_ebda_region();
 
-    /* could be an AMD 768MPX chipset. Reserve a page  before VGA to prevent
-       PCI prefetch into it (errata #56). Usually the page is reserved anyways,
-       unless you have no PS/2 mouse plugged in. */
+	/*
+	 * could be an AMD 768MPX chipset. Reserve a page before VGA to
+	 * prevent PCI prefetch into it (errata #56). Usually the page is
+	 * reserved anyways, unless you have no PS/2 mouse plugged in.
+	 */
 	if (boot_cpu_data.x86_vendor == X86_VENDOR_AMD &&
 	    boot_cpu_data.x86 == 6)
-	     reserve_bootmem(0xa0000 - 4096, 4096);
+		reserve_bootmem(0xa0000 - PAGE_SIZE, PAGE_SIZE);
 
 #ifdef CONFIG_SMP
 	/*
#1 lore raw
Quoted 3 lines · Andy Whitcroft

Convert the AMD 768MPX errata #56 fix to use PAGE_SIZE instead of using 4096 in line with other declarations in this file. Also take the oppotunity to match indentation.

Even if we use a different software page size in some future x86 release the workaround is to reserve 4096 bytes. I don't think this change makes sense therefore.

author
#2 lore raw
Quoted 3 lines

Even if we use a different software page size in some future x86 release the workaround is to reserve 4096 bytes. I don't think this change makes sense therefore.

I have always assumed that PAGE_SIZE in the bootmem allocator was hardware page size; such that if software PAGE_SIZE was to change that it would necessarily be replaced with two different constants. But fair enough. In that case perhaps we should at least fix the code indent etc. Revised patch below.

-apw

=== 8< ===
Fix indentation and layout of AMD 768MPX errata #56 fix.

Revision: $Rev: 620 $

Signed-off-by Andy Whitcroft <[email protected]>

diffstat 020-tidy-AMD-768MPX-fix

---
 setup.c |   10 ++++++----
 1 files changed, 6 insertions(+), 4 deletions(-)
arch/i386/kernel/setup.c +6 -4
diff -X /home/apw/brief/lib/vdiff.excl -rupN reference/arch/i386/kernel/setup.c current/arch/i386/kernel/setup.c
@@ -1052,12 +1052,14 @@ static unsigned long __init setup_memory
 	/* reserve EBDA region, it's a 4K region */
 	reserve_ebda_region();
 
-    /* could be an AMD 768MPX chipset. Reserve a page  before VGA to prevent
-       PCI prefetch into it (errata #56). Usually the page is reserved anyways,
-       unless you have no PS/2 mouse plugged in. */
+	/*
+	 * could be an AMD 768MPX chipset. Reserve a page before VGA to
+	 * prevent PCI prefetch into it (errata #56). Usually the page is
+	 * reserved anyways, unless you have no PS/2 mouse plugged in.
+	 */
 	if (boot_cpu_data.x86_vendor == X86_VENDOR_AMD &&
 	    boot_cpu_data.x86 == 6)
-	     reserve_bootmem(0xa0000 - 4096, 4096);
+		reserve_bootmem(0xa0000 - 4096, 4096);
 
 #ifdef CONFIG_SMP
 	/*
#3 lore raw
Quoted 2 lines · Andy Whitcroft

In that case perhaps we should at least fix the code indent
etc. Revised patch below.

If it's only a cosmetic patch anyway, it seems silly to do it without also fixing the misspelling of 'erratum' :)

Signature · 1 line
dwmw2