Hi Marcelo, here's a bunch of ACPI updates.
bk pull http://linux-acpi.bkbits.net/linux-acpi-2.4
This includes a rework of the ACPI config and cmdline options. It now
supports DMI-based blacklisting, and cmdline options have been
simplified to "acpi=off", "acpi=ht" (use ACPI for CPU enum only) and
"acpi=force" (to override the blacklist.)
It also includes some PCI IRQ routing changes, that seem to help some
people's systems work better.
Please apply.
Thanks -- Regards -- Andy
This will update the following files:
arch/i386/kernel/acpitable.c | 553 ------------------------
arch/i386/kernel/acpitable.h | 260 -----------
Documentation/Configure.help | 41 -
Documentation/kernel-parameters.txt | 18
Makefile | 2
arch/i386/kernel/Makefile | 1
arch/i386/kernel/acpi.c | 50 +-
arch/i386/kernel/dmi_scan.c | 394 ++++++++++++++---
arch/i386/kernel/io_apic.c | 14
arch/i386/kernel/mpparse.c | 37 +
arch/i386/kernel/setup.c | 99 ++--
arch/i386/kernel/smpboot.c | 2
drivers/Makefile | 2
drivers/acpi/Config.in | 68 +-
drivers/acpi/Makefile | 7
drivers/acpi/bus.c | 2
drivers/acpi/executer/exutils.c | 2
drivers/acpi/hardware/hwregs.c | 21
drivers/acpi/osl.c | 26 -
drivers/acpi/pci_irq.c | 15
drivers/acpi/pci_link.c | 100 ++--
drivers/acpi/processor.c | 1
drivers/acpi/tables.c | 120 ++---
drivers/acpi/tables/tbconvrt.c | 6
drivers/acpi/tables/tbget.c | 4
drivers/acpi/tables/tbinstal.c | 42 +
drivers/acpi/tables/tbrsdt.c | 2
drivers/acpi/tables/tbxfroot.c | 6
drivers/acpi/toshiba_acpi.c | 3
drivers/acpi/utilities/utglobal.c | 6
include/acpi/acconfig.h | 2
include/acpi/acpi_drivers.h | 2
include/acpi/platform/acenv.h | 6
include/asm-i386/acpi.h | 26 -
include/asm-i386/io_apic.h | 2
init/do_mounts.c | 12
36 files changed, 779 insertions(+), 1175 deletions(-)
through these ChangeSets:
<[email protected]> (03/07/29 1.1027)
ACPI: Allow irqs > 15 to use interrupt semantics other than PCI's
standard
active-low, level trigger. Make other changes as required for this.
(Andrew de
Quincey)
<[email protected]> (03/07/29 1.1026)
ACPI: Trivial changes that don't deserve their own changeset
<[email protected]> (03/07/29 1.1025)
ACPI: If notify handler fails to be removed properly, don't just
return, but
clean up other resources too (Daniele Bellucci)
<[email protected]> (03/07/29 1.1018.1.3)
minor cleanup of previous cset
<[email protected]> (03/07/29 1.1018.1.1)
rename CONFIG_ACPI_BOOT to CONFIG_ACPI_HT; "acpi=boot" to "acpi=ht"
Allow acpi=ht, or CONFIG_ACPI_HT to enable HT on systems w/o an MPS
table.
<[email protected]> (03/07/28 1.1022)
ACPI: toshiba_acpi update (John Belmonte)
<[email protected]> (03/07/28 1.1021)
ACPI: Better blacklist messages (Jasper Spaans)
<lenb@dhcppc3.> (03/07/23 1.1016.1.1)
ACPI: update acpi=boot code post testing
<lenb@dhcppc3.> (03/07/22 1.1003.38.2)
CONFIG_ACPI_BOOT is now a sub-set of CONFIG_ACPI. It enables just
boot-time config via ACPI --
useful for enabling HT w/o loading the AML interpreter.
When full CONFIG_ACPI is selected, cmdline "acpi=boot" runs the same
sub-set.
deleted acpitable.c, deleted broken "noht" cmdline option.
<[email protected]> (03/07/17 1.1015)
ACPI: fix intr enable for IA64 (davidm)
<[email protected]> (03/07/17 1.1014)
ACPI: Fix spinlock warnings (Bjorn Helgaas)
<lenb@dhcppc3.> (03/07/15 1.1002.2.3)
ACPI: DMI blacklist from UnitedLinux, plus "acpi=oldboot" scheme,
re-named "acpi=cpu"
<[email protected]> (03/07/14 1.1003.32.6)
ACPI: Dynamically allocate SDT list (suggested by Andi Kleen)
<[email protected]> (03/07/13 1.1003.32.5)
ACPI: Parse SSDTs in order discovered, as opposed to reverse order
(Hrvoje Habjanic)
<[email protected]> (03/07/13 1.1003.32.4)
ACPI: Fixes from FreeBSD and NetBSD. (Frank van der Linden, Thomas
Klausner,
Nate Lawson)
<[email protected]> (03/07/13 1.1003.32.3)
ACPI: DTRT regarding the watchdog during long stalls (Andrew Morton)
<[email protected]> (03/07/13 1.1002.7.2)
ACPI: Make things compile on gcc 3.3
<[email protected]> (03/07/13 1.1002.7.1)
ACPI: Make mp_bus_id_to_pci_bus initialization moderately more
correct
<lenb@dhcppc6.> (03/06/26 1.1002.2.2)
update kernel-parameters.txt with current acpi and acpismp
parameters.
-----------------------------
Andrew Grover
Intel Labs / Mobile Architecture
[email protected]
Andy,
Do you plan to accept below patch, H,J,L need this patch to get current 2.4 kernel boot on bigsur.
I think it also shuld be applied to 2.6 kernel.
Thanks,
Luming
===== drivers/acpi/osl.c 1.27 vs edited =====
--- 1.27/drivers/acpi/osl.c Thu Jul 10 04:20:31 2003
+++ edited/drivers/acpi/osl.c Wed Aug 6 16:16:38 2003
@@ -251,7 +251,12 @@
irq = acpi_fadt.sci_int;
#ifdef CONFIG_IA64
- irq = gsi_to_vector(irq);
+ irq = acpi_irq_to_vector(irq);
+ if (irq < 0) {
+ printk(KERN_ERR PREFIX "SCI (ACPI interrupt %d) not
registerd\n",
+ acpi_fadt.sci_int);
+ return AE_OK;
+ }
#endif
acpi_irq_irq = irq;
acpi_irq_handler = handler;
@@ -269,7 +274,7 @@
{
if (acpi_irq_handler) {
#ifdef CONFIG_IA64
- irq = gsi_to_vector(irq);
+ irq = acpi_irq_to_vector(irq);
#endif
free_irq(irq, acpi_irq);
acpi_irq_handler = NULL;
-----Original Message-----
From: Grover, Andrew
Sent: 2003?8?13? 6:36
To: Marcelo Tosatti
Cc: [email protected]; [email protected]
Subject: [BK PATCH] ACPI Updates for 2.4
Hi Marcelo, here's a bunch of ACPI updates.
bk pull http://linux-acpi.bkbits.net/linux-acpi-2.4
This includes a rework of the ACPI config and cmdline options. It now
supports DMI-based blacklisting, and cmdline options have been
simplified to "acpi=off", "acpi=ht" (use ACPI for CPU enum only) and
"acpi=force" (to override the blacklist.)
It also includes some PCI IRQ routing changes, that seem to help some
people's systems work better.
Please apply.
Thanks -- Regards -- Andy
This will update the following files:
arch/i386/kernel/acpitable.c | 553 ------------------------
arch/i386/kernel/acpitable.h | 260 -----------
Documentation/Configure.help | 41 -
Documentation/kernel-parameters.txt | 18
Makefile | 2
arch/i386/kernel/Makefile | 1
arch/i386/kernel/acpi.c | 50 +-
arch/i386/kernel/dmi_scan.c | 394 ++++++++++++++---
arch/i386/kernel/io_apic.c | 14
arch/i386/kernel/mpparse.c | 37 +
arch/i386/kernel/setup.c | 99 ++--
arch/i386/kernel/smpboot.c | 2
drivers/Makefile | 2
drivers/acpi/Config.in | 68 +-
drivers/acpi/Makefile | 7
drivers/acpi/bus.c | 2
drivers/acpi/executer/exutils.c | 2
drivers/acpi/hardware/hwregs.c | 21
drivers/acpi/osl.c | 26 -
drivers/acpi/pci_irq.c | 15
drivers/acpi/pci_link.c | 100 ++--
drivers/acpi/processor.c | 1
drivers/acpi/tables.c | 120 ++---
drivers/acpi/tables/tbconvrt.c | 6
drivers/acpi/tables/tbget.c | 4
drivers/acpi/tables/tbinstal.c | 42 +
drivers/acpi/tables/tbrsdt.c | 2
drivers/acpi/tables/tbxfroot.c | 6
drivers/acpi/toshiba_acpi.c | 3
drivers/acpi/utilities/utglobal.c | 6
include/acpi/acconfig.h | 2
include/acpi/acpi_drivers.h | 2
include/acpi/platform/acenv.h | 6
include/asm-i386/acpi.h | 26 -
include/asm-i386/io_apic.h | 2
init/do_mounts.c | 12
36 files changed, 779 insertions(+), 1175 deletions(-)
through these ChangeSets:
<[email protected]> (03/07/29 1.1027)
ACPI: Allow irqs > 15 to use interrupt semantics other than PCI's
standard
active-low, level trigger. Make other changes as required for this.
(Andrew de
Quincey)
<[email protected]> (03/07/29 1.1026)
ACPI: Trivial changes that don't deserve their own changeset
<[email protected]> (03/07/29 1.1025)
ACPI: If notify handler fails to be removed properly, don't just
return, but
clean up other resources too (Daniele Bellucci)
<[email protected]> (03/07/29 1.1018.1.3)
minor cleanup of previous cset
<[email protected]> (03/07/29 1.1018.1.1)
rename CONFIG_ACPI_BOOT to CONFIG_ACPI_HT; "acpi=boot" to "acpi=ht"
Allow acpi=ht, or CONFIG_ACPI_HT to enable HT on systems w/o an MPS
table.
<[email protected]> (03/07/28 1.1022)
ACPI: toshiba_acpi update (John Belmonte)
<[email protected]> (03/07/28 1.1021)
ACPI: Better blacklist messages (Jasper Spaans)
<lenb@dhcppc3.> (03/07/23 1.1016.1.1)
ACPI: update acpi=boot code post testing
<lenb@dhcppc3.> (03/07/22 1.1003.38.2)
CONFIG_ACPI_BOOT is now a sub-set of CONFIG_ACPI. It enables just
boot-time config via ACPI --
useful for enabling HT w/o loading the AML interpreter.
When full CONFIG_ACPI is selected, cmdline "acpi=boot" runs the same
sub-set.
deleted acpitable.c, deleted broken "noht" cmdline option.
<[email protected]> (03/07/17 1.1015)
ACPI: fix intr enable for IA64 (davidm)
<[email protected]> (03/07/17 1.1014)
ACPI: Fix spinlock warnings (Bjorn Helgaas)
<lenb@dhcppc3.> (03/07/15 1.1002.2.3)
ACPI: DMI blacklist from UnitedLinux, plus "acpi=oldboot" scheme,
re-named "acpi=cpu"
<[email protected]> (03/07/14 1.1003.32.6)
ACPI: Dynamically allocate SDT list (suggested by Andi Kleen)
<[email protected]> (03/07/13 1.1003.32.5)
ACPI: Parse SSDTs in order discovered, as opposed to reverse order
(Hrvoje Habjanic)
<[email protected]> (03/07/13 1.1003.32.4)
ACPI: Fixes from FreeBSD and NetBSD. (Frank van der Linden, Thomas
Klausner,
Nate Lawson)
<[email protected]> (03/07/13 1.1003.32.3)
ACPI: DTRT regarding the watchdog during long stalls (Andrew Morton)
<[email protected]> (03/07/13 1.1002.7.2)
ACPI: Make things compile on gcc 3.3
<[email protected]> (03/07/13 1.1002.7.1)
ACPI: Make mp_bus_id_to_pci_bus initialization moderately more
correct
<lenb@dhcppc6.> (03/06/26 1.1002.2.2)
update kernel-parameters.txt with current acpi and acpismp
parameters.
-----------------------------
Andrew Grover
Intel Labs / Mobile Architecture
[email protected]
On Maw, 2003-08-12 at 23:36, Grover, Andrew wrote:
> Hi Marcelo, here's a bunch of ACPI updates.
>
> bk pull http://linux-acpi.bkbits.net/linux-acpi-2.4
Is this available as a patch file somewhere ?
On Wed, Aug 13, 2003 at 02:54:58PM +0100, Alan Cox wrote:
> On Maw, 2003-08-12 at 23:36, Grover, Andrew wrote:
> > Hi Marcelo, here's a bunch of ACPI updates.
> >
> > bk pull http://linux-acpi.bkbits.net/linux-acpi-2.4
>
> Is this available as a patch file somewhere ?
Given that the last entry in the changelog dates 2003/07/29, I suppose that
the patch 20030730 from http://sf.net/projects/acpi/ is the same.
Cheers,
Willy
This ACPI update also fixes the irq routing issues I was seeing in
2.4.22-rc2. Previously, on Intel ICH5, I had to either disable
CONFIG_ACPI (which makes my 2nd HT CPU disappear) and boot with
'noapic', or disable CONFIG_SMP and boot with pci=noacpi. If I did not
do this, interrupts to the PCI slots would never be delivered.
With this update, 2.4 now matches 2.6 behavior:
full hyperthread, and interrupts routed correctly.
Jeff