Venki Pallipadi [Wed, 20 Jun 2007 21:24:52 +0000 (14:24 -0700)]
[CPUFREQ] ondemand: add a check to avoid negative load calculation
Due to rounding and inexact jiffy accounting, idle_ticks can sometimes
be higher than total_ticks. Make sure those cases are handled as
zero load case.
Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com> Signed-off-by: Dave Jones <davej@redhat.com>
Venki Pallipadi [Wed, 20 Jun 2007 21:24:00 +0000 (14:24 -0700)]
[CPUFREQ] Keep userspace governor quiet when it is not being used
Userspace governor registers a frequency change notifier at init time, even
when no CPU is set to userspace governor. Make it register only when
atleast one CPU is using userspace.
Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com> Signed-off-by: Dave Jones <davej@redhat.com>
Joshua Hoblitt [Tue, 22 May 2007 02:47:43 +0000 (16:47 -1000)]
[CPUFREQ] Kconfig powernow-k8 driver should depend on ACPI P-States driver
powernow-k8 really needs to use ACPI to function on SMP systems.
The current Kconfig allows us to build kernels which fail mysteriously
for some users due to us trying to automatically enable this, and
getting it wrong. It's easier to just present this as an option
to the user.
Signed-off-by: Joshua Hoblitt <jhoblitt@ifa.hawaii.edu> Signed-off-by: Dave Jones <davej@redhat.com>
Rafał Bilski [Tue, 5 Jun 2007 20:08:50 +0000 (22:08 +0200)]
[CPUFREQ] Longhaul - Replace ACPI functions with direct I/O
Current version of "bm status" bit test works as long as
no USB device is in use. When USB device is plugged in ACPI
function in this context is always returning 1. Until reboot.
Direct I/O is working fine even when many USB devices are
connected.
Change bm_timeout value to less annoying. 1000 is still much
more then worst case observed and it is much better when status
bit gets stuck.
Signed-off-by: Rafal Bilski <rafalbilski@interia.pl> Signed-off-by: Dave Jones <davej@redhat.com>
Rafał Bilski [Mon, 28 May 2007 19:58:09 +0000 (21:58 +0200)]
[CPUFREQ] Longhaul - Remove duplicate multipliers
Remove duplicate multipliers in clock_ratio table. On 1,4GHz
Nehemiah two frequencies are present twice in table. It isn't
fatal, but with voltage scaling enabled each will be set twice.
Signed-off-by: Rafal Bilski <rafalbilski@interia.pl> Signed-off-by: Dave Jones <davej@redhat.com>
Rafał Bilski [Mon, 28 May 2007 19:56:19 +0000 (21:56 +0200)]
[CPUFREQ] Longhaul - Embedded "conservative"
Longhaul with voltage scaling enabled works great on Ezra
CPU (Longhaul ver. 2). As long as "conservative" governor is
used. Both "ondemand" and "userspace" can change voltage
from min to max at once. Motherboard unfortunatly turns off
when vid difference is big. Longhaul was printing warning
message, but it is not enough. Now driver will have
"conservative" governor built in and will split bigger
changes to smaller ones.
Signed-off-by: Rafal Bilski <rafalbilski@interia.pl> Signed-off-by: Dave Jones <davej@redhat.com>
Venki Pallipadi [Wed, 23 May 2007 22:42:13 +0000 (15:42 -0700)]
[CPUFREQ] acpi-cpufreq: Proper ReadModifyWrite of PERF_CTL MSR
During recent acpi-cpufreq changes, writing to PERF_CTL msr
changed from RMW of entire 64 bit to RMW of low 32 bit and clearing of
upper 32 bit. Fix it back to do a proper RMW of the MSR.
Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com> Signed-off-by: Dave Jones <davej@redhat.com>
Rafał Bilski [Thu, 17 May 2007 20:39:02 +0000 (22:39 +0200)]
[CPUFREQ] Longhaul - Check ACPI "BM DMA in progress" bit
It is good idea to wait for PCI bus to become idle before
frequency change. Thanks to ACPI it is possible. It makes
sense only when northbridge support is in use because it is
only case in which we can disable arbiter after check if PCI
bus is busy.
Signed-off-by: Rafal Bilski <rafalbilski@interia.pl> Signed-off-by: Dave Jones <davej@redhat.com>
Rafał Bilski [Thu, 17 May 2007 20:33:46 +0000 (22:33 +0200)]
[CPUFREQ] Longhaul - Use all kinds of support
This patch is removing southbridge support as separate
kind of support. Instead it is used to make other kinds
of support more stable. Also northbridge and ACPI C3
support both will be used if both are available.
Signed-off-by: Rafal Bilski <rafalbilski@interia.pl> Signed-off-by: Dave Jones <davej@redhat.com>
Dave Jones [Mon, 14 May 2007 22:27:29 +0000 (18:27 -0400)]
[CPUFREQ] Correct revision mask for powernow-k8
Mark Langsdorf points out that the correct define for this
revision bump is 0x80000. Also to save us having to keep
renaming the #define, give it a more meaningful name.
Daniel Drake [Wed, 2 May 2007 22:19:05 +0000 (23:19 +0100)]
[CPUFREQ] powernow-k7: fix MHz rounding issue with perflib
When the PST tables are broken, powernow-k7 uses ACPI's processor_perflib to
deduce the available frequency multipliers from the _PSS tables.
Upon frequency change, processor_perflib performs some verification on the
frequency (checks that it's within allowable bounds).
powernow-k7 deals with absolute frequencies in KHz, whereas perflib only
deals with MHz values. When performing the above verification, perflib
multiplies the MHz values by 1000 to obtain the KHz value.
We then end up with situations like the following:
- powernow-k7 multiplies the multiplier by the FSB, and obtains a value
such as 1266768 KHz
- perflib belives the same state has frequency of 1266 MHz
- acpi_processor_ppc_notifier calls cpufreq_verify_within_limits to verify
that 1266768 is in the allowable range of 0 to 1266000 (i.e. 1266 * 1000)
- it's not, so that frequency is rejected
- the maximum CPU frequency is not reachable
This patch solves the problem by rounding up the MHz values stored in perflib's
tables. Additionally it corrects a broken URL.
It also fixes http://bugzilla.kernel.org/show_bug.cgi?id=8255 although this
case is a bit different: the frequencies in the _PSS tables are wildly wrong,
but we get better results if we force ACPI to respect the fsb * multiplier
calculations (even though it seems that the multiplier values aren't entirely
correct either).
Signed-off-by: Daniel Drake <dsd@gentoo.org> Signed-off-by: Dave Jones <davej@redhat.com>
SLUB cannot run on i386 at this point because i386 uses the page->private and
page->index field of slab pages for the pgd cache.
Make SLUB run on i386 by replacing the pgd slab cache with a quicklist.
Limit the changes as much as possible. Leave the improvised linked list in place
etc etc. This has been working here for a couple of weeks now.
Acked-by: William Lee Irwin III <wli@holomorphy.com> Signed-off-by: Christoph Lameter <clameter@sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Christian Krafft [Sat, 12 May 2007 17:37:01 +0000 (10:37 -0700)]
IPMI: Add PPC openfirmware unregister
When trying to load the ipmi_si module on a powerpc with no BMC (baseboard
management controller) the driver failes to load correctly, but doesn't
unregister itself from of_platform. So, on a second modprobe the kernel
crashes. This patch adds the missing unregister call.
Signed-off-by: Christian Krafft <krafft@de.ibm.com> Signed-off-by: Corey Minyard <minyard@acm.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Paul Fulghum [Sat, 12 May 2007 17:36:55 +0000 (10:36 -0700)]
tty: flush flip buffer on ldisc input queue flush
Flush the tty flip buffer when the line discipline input queue is flushed,
including the user call tcflush(TCIFLUSH/TCIOFLUSH). This prevents
unexpected stale data after a user application calls tcflush().
Signed-off-by: Alan Cox <alan@redhat.com> Cc: Antonino Ingargiola <tritemio@gmail.com> Signed-off-by: Paul Fulghum <paulkf@microgate.com> Cc: Theodore Ts'o <tytso@mit.edu> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Andi Kleen [Sat, 12 May 2007 15:27:40 +0000 (17:27 +0200)]
x86_64: Add asm/mtrr.h include for some builds
The earlier change to call the bp mtrr init from bugs.c broke
on some configurations due to missing includes. Noticed
by "Avuton Olrich" <avuton@gmail.com>
Dale Farnsworth [Sat, 12 May 2007 00:58:43 +0000 (10:58 +1000)]
[POWERPC] Add arch/powerpc support for the Motorola PrPMC2800
This finally adds the PPC_PRPMC2800 Kconfig option, the board setup
code (the setup and reset functions) and the defconfig, to support the
Motorola PrPMC2800 platform.
Signed-off-by: Dale Farnsworth <dale@farnsworth.org> Signed-off-by: Paul Mackerras <paulus@samba.org>
Mark A. Greer [Sat, 12 May 2007 00:58:18 +0000 (10:58 +1000)]
[POWERPC] Add bootwrapper support for Motorola PrPMC2800 platform
The Motorola PrPMC280 and PrPMC2800 processor modules sit on an F101 or
PrPMC2800 baseboard, respectively. There are several variants of each
type of processor module which can have different amounts of memory,
amounts of FLASH, cpu frequencies, and an mv64360 or an mv64362.
The bootwrapper code for that platform reads VPD from an I2C EEPROM
to determine the processor module variant. From the variant, the
amount of memory, etc. is determined and the device tree is updated
accordingly. If the variant cannot be determined (e.g., corrupted
VPD or a previously unknown variant), the property values already
in the device tree are used.
Also, the firmware for those platforms does not completely configure
the mv64x60 host bridge so that configuration is done here.
Signed-off-by: Mark A. Greer <mgreer@mvista.com> Signed-off-by: Paul Mackerras <paulus@samba.org>
Mark A. Greer [Sat, 12 May 2007 00:57:58 +0000 (10:57 +1000)]
[POWERPC] Add DTS file for the Motorola PrPMC2800 platform
Add the device tree source file for the prpmc2800 line of processor PMCs.
Several of the property values are updated by the bootwrapper but sane
defaults have been chosen in case the bootwrapper can't determine the
exact processor board variant. The defaults should allow the kernel
to boot despite having non-optimal device tree property values.
Signed-off-by: Mark A. Greer <mgreer@mvista.com> Signed-off-by: Paul Mackerras <paulus@samba.org>
Dale Farnsworth [Sat, 12 May 2007 00:57:35 +0000 (10:57 +1000)]
[POWERPC] Check cache coherency of kernel vs firmware
check_cache_coherency() verifies that the cache coherency setting of
the kernel (CONFIG_NOT_COHERENT_CACHE) matches that left by the firmware,
as indicated by coherency-off device tree property.
Signed-off-by: Dale Farnsworth <dale@farnsworth.org> Signed-off-by: Paul Mackerras <paulus@samba.org>
Dale Farnsworth [Sat, 12 May 2007 00:57:12 +0000 (10:57 +1000)]
[POWERPC] Add Marvell mv64x60 PCI bridge support
This patch adds PCI bridge support for the Marvell mv64x60 chip.
We also provide the ability to read/write the mv64x60 hotswap
register via sysfs if the hs_reg_valid property is set in the
device tree.
This patch creates platform_device entries for the Marvell mv64x60
I2C ports, based on information contained in device tree.
This driver (like the other mv64x60 drivers) are unusual in that it
works on both the MIPS and PowerPC architectures. Because of that,
the drivers do not support the normal PowerPC of_platform_bus_type.
They support platform_bus_type instead.
This patch creates platform_device entries for the Marvell mv64x60
ethernet controller ports, based on information contained in the
device tree.
This driver (like the other mv64x60 drivers) are unusual in that it
works on both the MIPS and PowerPC architectures. Because of that,
the drivers do not support the normal PowerPC of_platform_bus_type.
They support platform_bus_type instead.
This patch creates platform_device entries for the Marvell mv64x60
MPSC (multi-protocol serial controller) ports, based on information
contained in the device tree.
This driver (like the other mv64x60 drivers) are unusual in that it
works on both the MIPS and PowerPC architectures. Because of that,
the drivers do not support the normal PowerPC of_platform_bus_type.
They support platform_bus_type instead.
Dale Farnsworth [Sat, 12 May 2007 00:55:24 +0000 (10:55 +1000)]
[POWERPC] Add interrupt support for Marvell mv64x60 chips
There are 3 interrupt groups each with its own status/mask registers.
We use a separate struct irq_chip for each interrupt group and handle
interrupts in two stages or levels: level 1 selects the appropriate
struct irq_chip, and level 2 selects individual interrupts within
that irq_chip.
Signed-off-by: Dale Farnsworth <dale@farnsworth.org> Signed-off-by: Paul Mackerras <paulus@samba.org>
Mark A. Greer [Sat, 12 May 2007 00:54:53 +0000 (10:54 +1000)]
[POWERPC] Add bootwrapper support for Marvell/mv64x60 I2C
Some platforms support a variety processor modules with no method of
determining which exact processor module is being used except by
examining Vital Product Data (VPD). The modules may have different
amounts of memory, clock frequencies, etc. so reading the VPD becomes
necessary to correctly set properties in the device tree before its
passed to the kernel.
Often the VPD is stored in I2C EEPROMs so an I2C driver becomes necessary.
This I2C driver is for the I2C controller that's embedded on the Marvel
mv64x60 line of host bridges.
Signed-off-by: Mark A. Greer <mgreer@mvista.com> Signed-off-by: Paul Mackerras <paulus@samba.org>
Mark A. Greer [Sat, 12 May 2007 00:54:31 +0000 (10:54 +1000)]
[POWERPC] Add bootwrapper support for Marvell MPSC
The bootwrapper requires a serial driver to allow cmdline editing
and information reporting on the console. This driver is required
by platforms that boot a zImage and use the MPSC for the console.
Signed-off-by: Mark A. Greer <mgreer@mvista.com> Signed-off-by: Paul Mackerras <paulus@samba.org>
Mark A. Greer [Sat, 12 May 2007 00:54:05 +0000 (10:54 +1000)]
[POWERPC] Add bootwrapper support for Marvell/mv64x60 hostbridge
The mv64x60 host bridge has many windows between its various components
(cpu, system memory, ethernet ctlr, MPSC, DMA ctlr, PCI MEM, PCI I/O).
Unfortunately, the firmware on some of mv64x60-based platforms do not
properly or completely configure those windows (e.g., MPSC->system memory
windows not configured or CPU->PCI MEM space not configured).
So, the missing configuration needs to be done in either the bootwrapper
or in the kernel. To keep the kernel as clean as possible, it is done
in the bootwrapper. Note that I/O controller configuration is NOT being
done, its only the windows to allow the I/O controllers and other components
to access memory, etc. that is being done--drivers assume that their
controllers can already access system memory).
Table of routines and the windows they configure:
mv64x60_config_ctlr_windows() ENET->System Memory
MPSC->System Memory
IDMA->System Memory
Mark A. Greer [Sat, 12 May 2007 00:53:36 +0000 (10:53 +1000)]
[POWERPC] Add Makefile rules to wrap dts file in zImage
Add 'zImage.dts' and 'zImage.dts_initrd' build rules that automatically
compile and wrap a dts file from arch/powerpc/boot/dts into the zImage file.
The resulting zImage will be arch/powerpc/boot/zImage.dts.<platform> and
arch/powerpc/boot/zImage.dts_initrd.<platform>, respectively.
Having separate rules allows the user to choose whether to include a device
tree--and which device tree--at build time. This is useful when one Makefile
target builds a zImage that runs on several platforms except for differing
device trees. By just setting CONFIG_DEVICE_TREE and running "make zImage.dts"
the exact zImage you want is built without Makefile bloat or manually running
the wrapper script.
The dts file is expected to be arch/powerpc/boot/dts/$(CONFIG_DEVICE_TREE)
Signed-off-by: Mark A. Greer <mgreer@mvista.com> Signed-off-by: Paul Mackerras <paulus@samba.org>
Scott Wood [Fri, 11 May 2007 17:52:03 +0000 (03:52 +1000)]
[POWERPC] U-boot passes the initrd as start/end, not start/size.
The boot wrapper platform init code on 83xx and 85xx using the cuboot
platform type was incorrectly assuming that u-boot supplied the size
of the initrd, whereas it actually supplies the end address. This
fixes it.
Signed-off-by: Scott Wood <scottwood@freescale.com> Signed-off-by: Paul Mackerras <paulus@samba.org>
will schmidt [Fri, 11 May 2007 13:34:16 +0000 (23:34 +1000)]
[POWERPC] Simplify smp_space_timers
Greatly simplify the function smp_space_timers.
The stolen time calculation (per comment within the code) doesn't need the
half-jiffy stagger any more. There isn't an issue with bouncing off global
locks, so we really shouldn't need any sort of staggering at all.
However, the last_jiffy value still needs to be set. This removes the
extra stagger logic, and just sets the values.
This change should benefit applications that rely on barrier
synchronization, and will help cut down OS jitter.
Boot tested across the board (G5,power3,power4,power5,970mp blade).
Signed-off-by: Will Schmidt <will_schmidt@vnet.ibm.com> Signed-off-by: Paul Mackerras <paulus@samba.org>
Stephen Rothwell [Fri, 11 May 2007 05:42:44 +0000 (15:42 +1000)]
[POWERPC] Trivial ps3 warning fixes
Fixes warnings:
arch/powerpc/platforms/ps3/system-bus.c: In function 'ps3_map_sg':
arch/powerpc/platforms/ps3/system-bus.c:278: warning: unused variable 'i'
arch/powerpc/platforms/ps3/system-bus.c:277: warning: unused variable 'dev'
arch/powerpc/platforms/ps3/setup.c:103: warning: 'prealloc' defined but not used
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au> Acked-by: Geoff Levand <geoffrey.levand@am.sony.com> Signed-off-by: Paul Mackerras <paulus@samba.org>
Stephen Rothwell [Fri, 11 May 2007 05:40:36 +0000 (15:40 +1000)]
[POWERPC] Add missed include
fixes:
arch/powerpc/platforms/pasemi/idle.c: In function 'pasemi_system_reset_exception':
arch/powerpc/platforms/pasemi/idle.c:55: warning: implicit declaration of function 'do_IRQ'
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by: Paul Mackerras <paulus@samba.org>
Linus Torvalds [Fri, 11 May 2007 23:03:03 +0000 (16:03 -0700)]
Merge branch 'upstream-linus' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/libata-dev
* 'upstream-linus' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/libata-dev:
pata_platform: don't use generic ata_port_start
Use menuconfig objects: libata
add the ATI SB700 SATA controller device id to AHCI pci table
Add the combined mode for ATI SB700
pata_pcmcia: recognize 2GB CompactFlash from Transcend
git-libata-all: sata_via build fix
libata-acpi: clean up parameters and misc stuff
libata-acpi: s/CONFIG_SATA_ACPI/CONFIG_ATA_ACPI/
libata: give devices one last chance even if recovery failed with -EINVAL
libata: fallback to the other IDENTIFY on device error, take#2
libata: ignore EH scheduling during initialization
libata: clean up SFF init mess
libata: implement libata.spindown_compat
libata: reimplement suspend/resume support using sdev->manage_start_stop
Linus Torvalds [Fri, 11 May 2007 23:02:05 +0000 (16:02 -0700)]
Merge branch 'upstream-linus' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/netdev-2.6
* 'upstream-linus' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/netdev-2.6: (35 commits)
Add support for the Davicom DM9161A PHY
sky2: only disable 88e8056 on some boards
sky2: 88e8071 support not ready
skge: crash on shutdown/suspend
sky2: fix oops on shutdown
mlx4: don't use deprecated IRQ flags
netxen_nic_main don't use deprecated irq flags
Use menuconfig objects II - netdev/wan
Use menuconfig objects II - netdev/pcmcia
Use menuconfig objects II - netdev/atm
Use menuconfig objects: netdev
Use menuconfig objects: PHY
spidernet: remove unnecessary accesses to phy
S2IO: Statistics for link up/down and memory allocated/freed
S2IO: statistics for memory allocation failuers
S2IO: getringparam ethtool option
[PATCH] libertas: 64-bit cleanups
[PATCH] libertas: sparse fixes
[PATCH] libertas: fix missing unlock in TX error path
[PATCH] libertas: make debugfs.c sparse-clean
...
David Miller [Fri, 11 May 2007 20:26:44 +0000 (13:26 -0700)]
Fix assertion failure with MSI on sparc64
Today's find is a triggered assertion in msi_free_irqs() when the system
doesn't support MSI, in which case arch_setup_msi_irqs() always returns
an error.
The problem is that when this happens we branch into msi_free_irqs(), to
which you added the following assertion loop:
Well, if arch_setup_msi_irqs() fails, entry->irq will be zero and
although that's never assigned to any normal devices we use that IRQ
number for the timer interrupt on sparc64 so this assertion triggers.
Better to test for zero before doing the irq_has_action() assertion
thing.
Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Olof Johansson [Thu, 10 May 2007 04:00:20 +0000 (23:00 -0500)]
pata_platform: don't use generic ata_port_start
I have a system where I have a simple IDE controller that sits on a
local bus without bus master dma capability, and thus no dma_mapping
ops defined for the device/bus.
pata_platform works great for me, with the exception of using the generic
ata_port_start which tries to do a dmam_alloc_coherent.
Looks like it doesn't need to allocate a prd table at all, so replace it
with a dummy function instead.
Signed-off-by: Olof Johansson <olof@lixom.net> Signed-off-by: Jeff Garzik <jeff@garzik.org>
Jan Engelhardt [Fri, 11 May 2007 05:48:54 +0000 (22:48 -0700)]
Use menuconfig objects: libata
Use menuconfigs instead of menus, so the whole menu can be disabled at once
instead of going through all options.
Signed-off-by: Jan Engelhardt <jengelh@gmx.de> Cc: Jeff Garzik <jeff@garzik.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
Henry Su [Fri, 11 May 2007 05:48:51 +0000 (22:48 -0700)]
add the ATI SB700 SATA controller device id to AHCI pci table
Add the device ID to AHCI pci table for ATI SB700 SATA controller, the
subsequent chipset of SB600.
Signed-off-by: henry su<henry.su@amd.com> Cc: Jeff Garzik <jeff@garzik.org> Cc: Tejun Heo <htejun@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
Henry Su [Fri, 11 May 2007 05:48:50 +0000 (22:48 -0700)]
Add the combined mode for ATI SB700
Besides those modes in ATI SB600 SATA controller, ATI SB700 supports one
more mode:the combined mode.
The combined mode is a Legacy IDE mode used for compatibility with some old
OS without AHCI driver, but now it is not necessary for Linux since the
kernel has supported AHCI.
Signed-off-by: Luugi Marsan <luugi.marsan@amd.com> Cc: Jeff Garzik <jeff@garzik.org> Cc: Tejun Heo <htejun@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
Tejun Heo [Fri, 11 May 2007 12:46:51 +0000 (14:46 +0200)]
libata: give devices one last chance even if recovery failed with -EINVAL
After certain errors, some devices report complete garbage on
IDENTIFY. This can cause ata_dev_read_id() to fail with -EINVAL
resulting in immediate disabling of the device. Give the device one
last chance after -EINVAL to allow recovery from such situations. As
-EINVAL is triggered very rarely, this shouldn't cause any noticeable
affect on more common error paths.
Signed-off-by: Tejun Heo <htejun@gmail.com> Cc: Harald Dunkel <harald.dunkel@t-online.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
Tejun Heo [Fri, 11 May 2007 12:35:29 +0000 (14:35 +0200)]
libata: fallback to the other IDENTIFY on device error, take#2
It seems the world isn't as frank as we thought and some devices lie
about who they are. Fallback to the other IDENTIFY if IDENTIFY is
aborted by the device. As this is the strategy used by IDE for a long
time, it shouldn't cause too much problem.
Signed-off-by: Tejun Heo <htejun@gmail.com> Cc: William Thompson <wt@electro-mechanical.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
Tejun Heo [Tue, 1 May 2007 09:50:15 +0000 (11:50 +0200)]
libata: ignore EH scheduling during initialization
libata enables SCSI host during ATA host activation which happens
after IRQ handler is registered and IRQ is enabled. All ATA ports are
in frozen state when IRQ is enabled but frozen ports may raise limited
number of IRQs after being frozen - IOW, ->freeze() is not responsible
for clearing pending IRQs. During normal operation, the IRQ handler
is responsible for clearing spurious IRQs on frozen ports and it
usually doesn't require any extra code.
Unfortunately, during host initialization, the IRQ handler can end up
scheduling EH for a port whose SCSI host isn't initialized yet. This
results in OOPS in the SCSI midlayer. This is relatively short window
and scheduling EH for probing is the first thing libata does after
initialization, so ignoring EH scheduling until initialization is
complete solves the problem nicely.
This problem was spotted by Berck E. Nash in the following thread.
http://thread.gmane.org/gmane.linux.kernel/519412
Signed-off-by: Tejun Heo <htejun@gmail.com> Cc: Berck E. Nash <flyboy@gmail.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
Tejun Heo [Fri, 4 May 2007 10:43:58 +0000 (12:43 +0200)]
libata: clean up SFF init mess
The intention of using port_mask in SFF init helpers was to eventually
support exoctic configurations such as combination of legacy and
native port on the same controller. This never became actually
necessary and the related code always has been subtly broken one way
or the other. Now that new init model is in place, there is no reason
to make common helpers capable of handling all corner cases. Exotic
cases can simply dealt within LLDs as necessary.
This patch removes port_mask handling in SFF init helpers. SFF init
helpers don't take n_ports argument and interpret it into port_mask
anymore. All information is carried via port_info. n_ports argument
is dropped and always two ports are allocated. LLD can tell SFF to
skip certain port by marking it dummy. Note that SFF code has been
treating unuvailable ports this way for a long time until recent
breakage fix from Linus and is consistent with how other drivers
handle with unavailable ports.
This fixes 1-port legacy host handling still broken after the recent
native mode fix and simplifies SFF init logic. The following changes
are made...
* ata_pci_init_native_host() and ata_init_legacy_host() both now try
to initialized whatever they can and mark failed ports dummy. They
return 0 if any port is successfully initialized.
* ata_pci_prepare_native_host() and ata_pci_init_one() now doesn't
take n_ports argument. All info should be specified via port_info
array. Always two ports are allocated.
* ata_pci_init_bmdma() exported to be used by LLDs in exotic cases.
* port_info handling in all LLDs are standardized - all port_info
arrays are const stack variable named ppi. Unless the second port
is different from the first, its port_info is specified as NULL
(tells libata that it's identical to the last non-NULL port_info).
* pata_hpt37x/hpt3x2n: don't modify static variable directly. Make an
on-stack copy instead as ata_piix does.
* pata_uli: It has 4 ports instead of 2. Don't use
ata_pci_prepare_native_host(). Allocate the host explicitly and use
init helpers. It's simple enough.
Signed-off-by: Tejun Heo <htejun@gmail.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
Tejun Heo [Fri, 4 May 2007 19:28:48 +0000 (21:28 +0200)]
libata: implement libata.spindown_compat
Now that libata uses sd->manage_start_stop, libata spins down disk on
shutdown. In an attempt to compensate libata's previous shortcoming,
some distros sync and spin down disks attached via libata in their
shutdown(8). Some disks spin back up just to spin down again on
STANDBYNOW1 if the command is issued when the disk is spun down, so
this double spinning down causes problem.
This patch implements module parameter libata.spindown_compat which,
when set to one (default value), prevents libata from spinning down
disks on shutdown thus avoiding double spinning down. Note that
libata spins down disks for suspend to mem and disk, so with
libata.spindown_compat set to one, disks should be properly spun down
in all cases without modifying shutdown(8).
shutdown(8) should be fixed eventually. Some drive do spin up on
SYNCHRONZE_CACHE even when their cache is clean. Those disks
currently spin up briefly when sd tries to shutdown the device and
then the machine powers off immediately, which can't be good for the
head. We can't skip SYNCHRONIZE_CACHE during shudown as it can be
dangerous data integrity-wise.
So, this spindown_compat parameter is already scheduled for removal by
the end of the next year and here's what shutdown(8) should do.
* Check whether /sys/modules/libata/parameters/spindown_compat
exists. If it does, write 0 to it.
* For each libata harddisk {
* Check whether /sys/class/scsi_disk/h:c:i:l/manage_start_stop
exists. Iff it doesn't, synchronize cache and spin the disk
down as before.
}
The above procedure will make shutdown(8) work properly with kernels
before this change, ones with this workaround and later ones without
it.
To accelerate shutdown(8) updates, if the compat mode is in use, this
patch prints BIG FAT warning for five seconds during shutdown (the
optimal interval to annoy the user just the right amount discovered by
hours of tireless usability testing).
Signed-off-by: Tejun Heo <htejun@gmail.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
Tejun Heo [Fri, 4 May 2007 19:27:47 +0000 (21:27 +0200)]
libata: reimplement suspend/resume support using sdev->manage_start_stop
Reimplement suspend/resume support using sdev->manage_start_stop.
* Device suspend/resume is now SCSI layer's responsibility and the
code is simplified a lot.
* DPM is dropped. This also simplifies code a lot. Suspend/resume
status is port-wide now.
* ata_scsi_device_suspend/resume() and ata_dev_ready() removed.
* Resume now has to wait for disk to spin up before proceeding. I
couldn't find easy way out as libata is in EH waiting for the
disk to be ready and sd is waiting for EH to complete to issue
START_STOP.
* sdev->manage_start_stop is set to 1 in ata_scsi_slave_config().
This fixes spindown on shutdown and suspend-to-disk.
Signed-off-by: Tejun Heo <htejun@gmail.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
Use DMI to add a blacklist of broken boards (so far only one).
Hopefully, the problems will be solved later, and the the whole
blacklist can disappear.
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
If device is not fails during module startup (like unsupported chip
version) then driver would crash dereferencing a null pointer, on shutdown
or suspend/resume.
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
If the device is fails during module startup for some reason like
unsupported chip version then the driver would crash dereferencing a
null pointer, on shutdown or suspend/resume.
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
Andrew Morton [Fri, 11 May 2007 05:53:00 +0000 (22:53 -0700)]
netxen_nic_main don't use deprecated irq flags
drivers/net/netxen/netxen_nic_main.c: In function 'netxen_nic_open':
drivers/net/netxen/netxen_nic_main.c:738: warning: 'deprecated_irq_flag' is deprecated (declared at include/linux/interrupt.h:66)
drivers/net/netxen/netxen_nic_main.c:738: warning: 'deprecated_irq_flag' is deprecated (declared at include/linux/interrupt.h:66)
Cc: Jeff Garzik <jeff@garzik.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
Jan Engelhardt [Fri, 11 May 2007 05:52:59 +0000 (22:52 -0700)]
Use menuconfig objects II - netdev/wan
Change Kconfig objects from "menu, config" into "menuconfig" so
that the user can disable the whole feature without having to
enter the menu first.
Signed-off-by: Jan Engelhardt <jengelh@gmx.de> Cc: Jeff Garzik <jeff@garzik.org> Cc: Krzysztof Halasa <khc@pm.waw.pl> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
Jan Engelhardt [Fri, 11 May 2007 05:52:58 +0000 (22:52 -0700)]
Use menuconfig objects II - netdev/pcmcia
Change Kconfig objects from "menu, config" into "menuconfig" so
that the user can disable the whole feature without having to
enter the menu first.
Signed-off-by: Jan Engelhardt <jengelh@gmx.de> Cc: Jeff Garzik <jeff@garzik.org> Cc: Dominik Brodowski <linux@dominikbrodowski.net> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
projects / linux-2.6 / log