RE: [PATCH] 2.4/2.6 use xdsdt to print table header

RE: [PATCH] 2.4/2.6 use xdsdt to print table header

[Yu, Luming]

>    I'm resending this patch to get it into the main ACPI source.  This
> fixes a problem where the DSDT pointer in the FADT is NULL because it
> uses the 64bit XDSDT instead.  The current code is happy to map a NULL
> address and return success to the caller.  This can crash the 
> system or
> printout garbage headers to the console.  It's a simple 
> matter to check
> table revision and use the XDSDT in favor of the DSDT.  This has been
> living happily in both the 2.4 and 2.6 ia64 tree for some 
> time.  Please
> accept.  Thanks,

I just checked with http://lia64.bkbits.net:8080/linux-ia64-2.4 .
The patch has been merged. Please take a look at
http://lia64.bkbits.net:8080/linux-ia64-2.4/diffs/drivers/acpi/tables.c@
1.11.1.1?nav=index.html|src/.|src/drivers|src/drivers/acpi|hist/drivers/
acpi/tables.c
and
http://lia64.bkbits.net:8080/linux-ia64-2.4/diffs/drivers/acpi/tables.c@
1.11.1.2?nav=index.html|src/.|src/drivers|src/drivers/acpi|hist/drivers/
acpi/tables.c


-------------------------------------------------------
This SF.net email is sponsored by: Perforce Software.
Perforce is the Fast Software Configuration Management System offering
advanced branching capabilities and atomic changes on 50+ platforms.
Free Eval! http://www.perforce.com/perforce/loadprog.html

RE: [PATCH] 2.4/2.6 use xdsdt to print table header

[Alex Williamson]

On Tue, 2004-01-13 at 00:38, Yu, Luming wrote:
> >    I'm resending this patch to get it into the main ACPI source.  This
> > fixes a problem where the DSDT pointer in the FADT is NULL because it
> > uses the 64bit XDSDT instead.  The current code is happy to map a NULL
> > address and return success to the caller.  This can crash the 
> > system or
> > printout garbage headers to the console.  It's a simple 
> > matter to check
> > table revision and use the XDSDT in favor of the DSDT.  This has been
> > living happily in both the 2.4 and 2.6 ia64 tree for some 
> > time.  Please
> > accept.  Thanks,
> 
> I just checked with http://lia64.bkbits.net:8080/linux-ia64-2.4 .
> The patch has been merged. Please take a look at

   Right, it's in the ia64 trees, but understandably the ia64 arch
maintainers don't want to carry this non-ia64 specific patch
indefinitely.  This is a generic ACPI issue for an system that chooses
to use the xdsdt in place of the dsdt.  It should be included in the
ACPI tree and pushed up from there.  Thanks,

	Alex

-- 
Alex Williamson                             HP Linux & Open Source Lab



-------------------------------------------------------
This SF.net email is sponsored by: Perforce Software.
Perforce is the Fast Software Configuration Management System offering
advanced branching capabilities and atomic changes on 50+ platforms.
Free Eval! http://www.perforce.com/perforce/loadprog.html

RE: [PATCH] 2.4/2.6 use xdsdt to print table header

[Yu, Luming]

> indefinitely.  This is a generic ACPI issue for an system that chooses
> to use the xdsdt in place of the dsdt.  It should be included in the
> ACPI tree and pushed up from there.  Thanks,
> 
All right, I'll push this to Len.


-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004
Premiere Conference on Open Tools Development and Integration
See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn

[PATCH] 2.4/2.6 use xdsdt to print table header

[Alex Williamson]

   I'm resending this patch to get it into the main ACPI source.  This
fixes a problem where the DSDT pointer in the FADT is NULL because it
uses the 64bit XDSDT instead.  The current code is happy to map a NULL
address and return success to the caller.  This can crash the system or
printout garbage headers to the console.  It's a simple matter to check
table revision and use the XDSDT in favor of the DSDT.  This has been
living happily in both the 2.4 and 2.6 ia64 tree for some time.  Please
accept.  Thanks,

	Alex

-- 
Alex Williamson                             HP Linux & Open Source Lab

diff -Nru a/drivers/acpi/tables.c b/drivers/acpi/tables.c
--- a/drivers/acpi/tables.c	Mon Jan 12 15:37:12 2004
+++ b/drivers/acpi/tables.c	Mon Jan 12 15:37:12 2004
@@ -262,10 +262,17 @@
 
 	/* Map the DSDT header via the pointer in the FADT */
 	if (id == ACPI_DSDT) {
-		struct acpi_table_fadt *fadt = (struct acpi_table_fadt *) *header;
+		struct fadt_descriptor_rev2 *fadt = (struct fadt_descriptor_rev2 *) *header;
+
+		if (fadt->revision == 3 && fadt->Xdsdt) {
+			*header = (void *) __acpi_map_table(fadt->Xdsdt,
+					sizeof(struct acpi_table_header));
+		} else if (fadt->V1_dsdt) {
+			*header = (void *) __acpi_map_table(fadt->V1_dsdt,
+					sizeof(struct acpi_table_header));
+		} else
+			*header = 0;
 
-		*header = (void *) __acpi_map_table(fadt->dsdt_addr,
-				sizeof(struct acpi_table_header));
 		if (!*header) {
 			printk(KERN_WARNING PREFIX "Unable to map DSDT\n");
 			return -ENODEV;





-------------------------------------------------------
This SF.net email is sponsored by: Perforce Software.
Perforce is the Fast Software Configuration Management System offering
advanced branching capabilities and atomic changes on 50+ platforms.
Free Eval! http://www.perforce.com/perforce/loadprog.html

Re: [PATCH] 2.4/2.6 use xdsdt to print table header

[Len Brown]

Accepted.

Thanks Alex,
-Len

On Mon, 2004-01-12 at 19:32, Alex Williamson wrote:
>    I'm resending this patch to get it into the main ACPI source.  This
> fixes a problem where the DSDT pointer in the FADT is NULL because it
> uses the 64bit XDSDT instead.  The current code is happy to map a NULL
> address and return success to the caller.  This can crash the system or
> printout garbage headers to the console.  It's a simple matter to check
> table revision and use the XDSDT in favor of the DSDT.  This has been
> living happily in both the 2.4 and 2.6 ia64 tree for some time.  Please
> accept.  Thanks,
> 
> 	Alex



-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004
Premiere Conference on Open Tools Development and Integration
See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn