We have a large university system in the UK that is experiencing very long
delays modprobing the driver for a specific I/O device. The delay is from
8-10 minutes per device and there are 31 devices in the system. This 4 to
5 hour delay in starting up those I/O devices is very much a burden on the
customer.
There are two causes for requiring a restart/reload of the drivers. First
is periodic preventive maintenance (PM) and the second is if any of the
devices experience a fatal error. Both of these trigger this excessively
long delay in bringing the system back up to full capability.
The problem was tracked down to a very slow IOREMAP operation and the
excessively long ioresource lookup to insure that the user is not
attempting to ioremap RAM. These patches provide a speed up to that
function.
The modprobe time appears to be affected quite a bit by previous activity
on the ioresource list, which I suspect is due to cache preloading. While
the overall improvement is impacted by other overhead of starting the
devices, this drastically improves the modprobe time.
Also our system is considerably smaller so the percentages gained will not
be the same. Best case improvement with the modprobe on our 20 device
smallish system was from 'real 5m51.913s' to 'real 0m18.275s'.
This patch (of 2):
Since the ioremap operation is verifying that the specified address range
is NOT RAM, it will search the entire ioresource list if the condition is
true. To make matters worse, it does this one 4k page at a time. For a
128M BAR region this is 32 passes to determine the entire region does not
contain any RAM addresses.
This patch provides another resource lookup function, region_is_ram, that
searches for the entire region specified, verifying that it is completely
contained within the resource region. If it is found, then it is checked
to be RAM or not, within a single pass.
The return result reflects if it was found or not (-1), and whether it is
RAM (1) or not (0). This allows the caller to fallback to the previous
page by page search if it was not found.
[akpm@linux-foundation.org: fix spellos and typos in comment]
Signed-off-by: Mike Travis <travis@sgi.com>
Acked-by: Alex Thorlton <athorlton@sgi.com>
Reviewed-by: Cliff Wickman <cpw@sgi.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Mark Salter <msalter@redhat.com>
Cc: Dave Young <dyoung@redhat.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Fix UV call into kgdb to depend only on whether KGDB is defined
and not both KGDB and KDB. This allows the power nmi command to
use the gdb remote connection if enabled. Note new action of
'kgdb' needs to be set as well to indicate user wants to wait
for gdb to be connected. If it's set to 'kdb' then an error
message is displayed if KDB is not configured.
Also note that if both KGDB and KDB are enabled, then the action
of 'kgdb' or 'kdb' has no affect on which is used. See the KGDB
documentation for further information.
Signed-off-by: Mike Travis <travis@sgi.com>
Reviewed-by: Hedi Berriche <hedi@sgi.com>
Cc: Russ Anderson <rja@sgi.com>
Cc: Jason Wessel <jason.wessel@windriver.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Link: http://lkml.kernel.org/r/20140114162551.635540667@asylum.americas.sgi.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
This patch restores the capability to enter KDB (and KGDB) from
the UV NMI handler. This is needed because the UV system
console is not capable of sending the 'break' signal to the
serial console port. It is also useful when the kernel is hung
in such a way that it isn't responding to normal external I/O,
so sending 'g' to sysreq-trigger does not work either.
Another benefit of the external NMI command is that all the cpus
receive the NMI signal at roughly the same time so they are more
closely aligned timewise.
It utilizes the newly added kgdb_nmicallin function to gain
entry to KGDB/KDB by the master. The slaves still enter via the
standard kgdb_nmicallback function. It also uses the new
'send_ready' pointer to tell KGDB/KDB to signal the slaves when
to proceed into the KGDB slave loop.
It is enabled when the nmi action is set to "kdb" and the kernel
is built with CONFIG_KDB enabled. Note that if kgdb is
connected that interface will be used instead.
Signed-off-by: Mike Travis <travis@sgi.com>
Reviewed-by: Dimitri Sivanich <sivanich@sgi.com>
Reviewed-by: Hedi Berriche <hedi@sgi.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Cc: Jason Wessel <jason.wessel@windriver.com>
Link: http://lkml.kernel.org/r/20131002151418.089692683@asylum.americas.sgi.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
This patch adds a kgdb_nmicallin() interface that can be used by
external NMI handlers to call the KGDB/KDB handler. The primary
need for this is for those types of NMI interrupts where all the
CPUs have already received the NMI signal. Therefore no
send_IPI(NMI) is required, and in fact it will cause a 2nd
unhandled NMI to occur. This generates the "Dazed and Confuzed"
messages.
Since all the CPUs are getting the NMI at roughly the same time,
it's not guaranteed that the first CPU that hits the NMI handler
will manage to enter KGDB and set the dbg_master_lock before the
slaves start entering. The new argument "send_ready" was added
for KGDB to signal the NMI handler to release the slave CPUs for
entry into KGDB.
Signed-off-by: Mike Travis <travis@sgi.com>
Acked-by: Jason Wessel <jason.wessel@windriver.com>
Reviewed-by: Dimitri Sivanich <sivanich@sgi.com>
Reviewed-by: Hedi Berriche <hedi@sgi.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Link: http://lkml.kernel.org/r/20131002151417.928886849@asylum.americas.sgi.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The current UV NMI handler has not been updated for the changes
in the system NMI handler and the perf operations. The UV NMI
handler reads an MMR in the UV Hub to check to see if the NMI
event was caused by the external 'system NMI' that the operator
can initiate on the System Mgmt Controller.
The problem arises when the perf tools are running, causing
millions of perf events per second on very large CPU count
systems. Previously this was okay because the perf NMI handler
ran at a higher priority on the NMI call chain and if the NMI
was a perf event, it would stop calling other NMI handlers
remaining on the NMI call chain.
Now the system NMI handler calls all the handlers on the NMI
call chain including the UV NMI handler. This causes the UV NMI
handler to read the MMRs at the same millions per second rate.
This can lead to significant performance loss and possible
system failures. It also can cause thousands of 'Dazed and
Confused' messages being sent to the system console. This
effectively makes perf tools unusable on UV systems.
To avoid this excessive overhead when perf tools are running,
this code has been optimized to minimize reading of the MMRs as
much as possible, by moving to the NMI_UNKNOWN notifier chain.
This chain is called only when all the users on the standard
NMI_LOCAL call chain have been called and none of them have
claimed this NMI.
There is an exception where the NMI_LOCAL notifier chain is
used. When the perf tools are in use, it's possible that the UV
NMI was captured by some other NMI handler and then either
ignored or mistakenly processed as a perf event. We set a
per_cpu ('ping') flag for those CPUs that ignored the initial
NMI, and then send them an IPI NMI signal. The NMI_LOCAL
handler on each cpu does not need to read the MMR, but instead
checks the in memory flag indicating it was pinged. There are
two module variables, 'ping_count' indicating how many requested
NMI events occurred, and 'ping_misses' indicating how many stray
NMI events. These most likely are perf events so it shows the
overhead of the perf NMI interrupts and how many MMR reads were avoided.
This patch also minimizes the reads of the MMRs by having the
first cpu entering the NMI handler on each node set a per HUB
in-memory atomic value. (Having a per HUB value avoids sending
lock traffic over NumaLink.) Both types of UV NMIs from the SMI
layer are supported.
Signed-off-by: Mike Travis <travis@sgi.com>
Reviewed-by: Dimitri Sivanich <sivanich@sgi.com>
Reviewed-by: Hedi Berriche <hedi@sgi.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Cc: Jason Wessel <jason.wessel@windriver.com>
Link: http://lkml.kernel.org/r/20130923212500.353547733@asylum.americas.sgi.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The audio driver mistakenly allows 64 bit addresses to be created for
the audio driver on Nvidia GPUs. Unfortunately, the hardware normally
only supports up to 40 bits of DMA. This can cause system panics as
well as misdirected data when the address is > 40 bits as the upper
part the address is truncated.
Signed-off-by: Mike Travis <travis@sgi.com>
Reviewed-by: Mike Habeck <habeck@sgi.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
This patch trims the MMR register definitions after the updates for the
SGI UV3 system have been applied. Note that because these definitions
are automatically generated from the RTL we cannot control the length
of the names. Therefore there are lines that exceed 80 characters.
Signed-off-by: Mike Travis <travis@sgi.com>
Link: http://lkml.kernel.org/r/20130211194509.173026880@gulag1.americas.sgi.com
Acked-by: Russ Anderson <rja@sgi.com>
Reviewed-by: Dimitri Sivanich <sivanich@sgi.com>
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
This patch adds support for the SGI UV3 hub to the common x2apic
functions. The primary changes are to account for the similarities
between UV2 and UV3 which are encompassed within the "UVX" nomenclature.
One significant difference within UV3 is the handling of the MMIOH
regions which are redirected to the target blade (with the device) in
a different manner. It also now has two MMIOH regions for both small and
large BARs. This aids in limiting the amount of physical address space
removed from real memory that's used for I/O in the max config of 64TB.
Signed-off-by: Mike Travis <travis@sgi.com>
Link: http://lkml.kernel.org/r/20130211194508.752924185@gulag1.americas.sgi.com
Acked-by: Russ Anderson <rja@sgi.com>
Reviewed-by: Dimitri Sivanich <sivanich@sgi.com>
Cc: Alexander Gordeev <agordeev@redhat.com>
Cc: Suresh Siddha <suresh.b.siddha@intel.com>
Cc: Michael S. Tsirkin <mst@redhat.com>
Cc: Steffen Persvold <sp@numascale.com>
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
This patch updates the MMR register definitions for the SGI UV3 system.
Note that because these definitions are automatically generated from
the RTL we cannot control the length of the names. Therefore there are
lines that exceed 80 characters.
All the new MMR definitions are added in this patch. The patches that
follow then update the references. The last patch is a "trim" patch
which reduces the size of the MMR definitions file by about a third.
This keeps "bi-sectability" in place as the intermediate patches would
not compile correctly if the trimmed MMR defines were done first.
Signed-off-by: Mike Travis <travis@sgi.com>
Link: http://lkml.kernel.org/r/20130211194508.326204556@gulag1.americas.sgi.com
Acked-by: Russ Anderson <rja@sgi.com>
Reviewed-by: Dimitri Sivanich <sivanich@sgi.com>
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
When using the 1:1 (identity) PCI DMA remapping, PCI Host Bridge devices
that do not use the IOMMU causes a kernel panic. Fix that by not
inserting those devices into the si_domain.
Signed-off-by: Mike Travis <travis@sgi.com>
Reviewed-by: Mike Habeck <habeck@sgi.com>
Cc: stable@kernel.org
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
The __intel_map_single function is not honoring the passed in DMA mask.
This results in not using the coherent DMA mask when called from
intel_alloc_coherent().
Signed-off-by: Mike Travis <travis@sgi.com>
Acked-by: Chris Wright <chrisw@sous-sol.org>
Reviewed-by: Mike Habeck <habeck@sgi.com>
Cc: stable@kernel.org
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
When there are a large count of PCI devices, and the pass through
option for iommu is set, much time is spent in the identity_mapping
function hunting though the iommu domains to check if a specific
device is "identity mapped".
Speed up the function by checking the cached info to see if
it's mapped to the static identity domain.
Signed-off-by: Mike Travis <travis@sgi.com>
Reviewed-by: Mike Habeck <habeck@sgi.com>
Cc: stable@kernel.org
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
On larger systems, because of the numerous ACPI, Bootmem and EFI messages,
the static log buffer overflows before the larger one specified by the
log_buf_len param is allocated. Minimize the overflow by allocating the
new log buffer as soon as possible.
On kernels without memblock, a later call to setup_log_buf from
kernel/init.c is the fallback.
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: fix CONFIG_PRINTK=n build]
Signed-off-by: Mike Travis <travis@sgi.com>
Cc: Yinghai Lu <yhlu.kernel@gmail.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Jack Steiner <steiner@sgi.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Manually adjusting the smp_affinity for IRQ's becomes unwieldy when the
cpu count is large.
Setting smp affinity to cpus 256 to 263 would be:
echo 000000ff,00000000,00000000,00000000,00000000,00000000,00000000,00000000 > smp_affinity
instead of:
echo 256-263 > smp_affinity_list
Think about what it looks like for cpus around say, 4088 to 4095.
We already have many alternate "list" interfaces:
/sys/devices/system/cpu/cpuX/indexY/shared_cpu_list
/sys/devices/system/cpu/cpuX/topology/thread_siblings_list
/sys/devices/system/cpu/cpuX/topology/core_siblings_list
/sys/devices/system/node/nodeX/cpulist
/sys/devices/pci***/***/local_cpulist
Add a companion interface, smp_affinity_list to use cpu lists instead of
cpu maps. This conforms to other companion interfaces where both a map
and a list interface exists.
This required adding a bitmap_parselist_user() function in a manner
similar to the bitmap_parse_user() function.
[akpm@linux-foundation.org: make __bitmap_parselist() static]
Signed-off-by: Mike Travis <travis@sgi.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Jack Steiner <steiner@sgi.com>
Cc: Lee Schermerhorn <lee.schermerhorn@hp.com>
Cc: Andy Shevchenko <andy.shevchenko@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Update pci_set_vga_state to call arch dependent functions to enable Legacy
VGA I/O transactions to be redirected to correct target.
[akpm@linux-foundation.org: make pci_register_set_vga_state() __init]
Signed-off-by: Mike Travis <travis@sgi.com>
LKML-Reference: <201002022238.o12McE1J018723@imap1.linux-foundation.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Robin Holt <holt@sgi.com>
Cc: Jack Steiner <steiner@sgi.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Jesse Barnes <jbarnes@virtuousgeek.org>
Cc: David Airlie <airlied@linux.ie>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
This fixes the problem of the initialization code not correctly
mapping the entire MMIO space on a UV system. A side effect is
the map_high() interface needed to be changed to accommodate
different address and size shifts.
Signed-off-by: Mike Travis <travis@sgi.com>
Reviewed-by: Mike Habeck <habeck@sgi.com>
Cc: <stable@kernel.org>
Cc: Jack Steiner <steiner@sgi.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
LKML-Reference: <4B479202.7080705@sgi.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
The wrong address was being used to write the SCIR led regs on remote
hubs. Also, there was an inconsistency between how BIOS and the kernel
indexed these regs. Standardize on using the lower 6 bits of the APIC
ID as the index.
This patch fixes the problem of writing to an errant address to a
cpu # >= 64.
Signed-off-by: Mike Travis <travis@sgi.com>
Reviewed-by: Jack Steiner <steiner@sgi.com>
Cc: stable@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The wrong address was being used to write the SCIR led regs on
remote hubs. Also, there was an inconsistency between how BIOS
and the kernel indexed these regs. Standardize on using the
lower 6 bits of the APIC ID as the index.
This patch fixes the problem of writing to an errant address to
a cpu # >= 64.
Signed-off-by: Mike Travis <travis@sgi.com>
Reviewed-by: Jack Steiner <steiner@sgi.com>
Cc: Robin Holt <holt@sgi.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: stable@kernel.org
LKML-Reference: <4B3922F9.3060905@sgi.com>
[ v2: fix a number of annoying checkpatch artifacts and whitespace noise ]
Signed-off-by: Ingo Molnar <mingo@elte.hu>
When there are a large number of processors in a system, there
is an excessive amount of messages sent to the system console.
It's estimated that with 4096 processors in a system, and the
console baudrate set to 56K, the startup messages will take
about 84 minutes to clear the serial port.
This set of patches limits the number of repetitious messages
which contain no additional information. Much of this information
is obtainable from the /proc and /sysfs. Some of the messages
are also sent to the kernel log buffer as KERN_DEBUG messages so
dmesg can be used to examine more closely any details specific to
a problem.
The new cpu bootup sequence for system_state == SYSTEM_BOOTING:
Booting Node 0, Processors #1#2#3#4#5#6#7 Ok.
Booting Node 1, Processors #8#9#10#11#12#13#14#15 Ok.
...
Booting Node 3, Processors #56#57#58#59#60#61#62#63 Ok.
Brought up 64 CPUs
After the system is running, a single line boot message is displayed
when CPU's are hotplugged on:
Booting Node %d Processor %d APIC 0x%x
Status of the following lines:
CPU: Physical Processor ID: printed once (for boot cpu)
CPU: Processor Core ID: printed once (for boot cpu)
CPU: Hyper-Threading is disabled printed once (for boot cpu)
CPU: Thermal monitoring enabled printed once (for boot cpu)
CPU %d/0x%x -> Node %d: removed
CPU %d is now offline: only if system_state == RUNNING
Initializing CPU#%d: KERN_DEBUG
Signed-off-by: Mike Travis <travis@sgi.com>
LKML-Reference: <4B219E28.8080601@sgi.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Impact: cleanup, update to new cpumask API
Irq_desc.affinity and irq_desc.pending_mask are now cpumask_var_t's
so access to them should be using the new cpumask API.
Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Kyle McMartin <kyle@mcmartin.ca>
Impact: use new work_on_cpu function to reduce stack usage
Replace the saving of current->cpus_allowed and set_cpus_allowed_ptr() with
a work_on_cpu function for drv_read() and drv_write().
Basically converts do_drv_{read,write} into "work_on_cpu" functions that
are now called by drv_read and drv_write.
Note: This patch basically reverts 50c668d6 which reverted 7503bfba, now
that the work_on_cpu() function is more stable.
Signed-off-by: Mike Travis <travis@sgi.com>
Acked-by: Rusty Russell <rusty@rustcorp.com.au>
Tested-by: Dieter Ries <clip2@gmx.de>
Tested-by: Maciej Rutecki <maciej.rutecki@gmail.com>
Cc: Dave Jones <davej@redhat.com>
Cc: <cpufreq@vger.kernel.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: add debug warning
Fire off one message if two apic's discovered with different
apic versions. (this code is only called during CPU init)
The goal of this is to pave the way of the removal of the apic_version[]
array. We dont expect any apic version incompatibilities in the x86
landscape of systems [if so we dont handle them very well and probably
never will handle deep apic version assymetries well], but it's prudent
to have a debug check for one kernel cycle nevertheless.
Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: use new work_on_cpu function to reduce stack usage
Replace the saving of current->cpus_allowed and set_cpus_allowed_ptr() with
a work_on_cpu function for drv_read() and drv_write().
Basically converts do_drv_{read,write} into "work_on_cpu" functions that
are now called by drv_read and drv_write.
Note: This patch basically reverts 50c668d6 which reverted 7503bfba, now
that the work_on_cpu() function is more stable.
Signed-off-by: Mike Travis <travis@sgi.com>
Acked-by: Rusty Russell <rusty@rustcorp.com.au>
Tested-by: Dieter Ries <clip2@gmx.de>
Tested-by: Maciej Rutecki <maciej.rutecki@gmail.com>
Cc: Dave Jones <davej@redhat.com>
Cc: <cpufreq@vger.kernel.org>
[ Based on original patch from Christoph Lameter and Mike Travis. ]
* Ruggedize some calls in setup_percpu.c to prevent mishaps
in early calls, particularly for non-critical functions.
* Cleanup DEBUG_PER_CPU_MAPS usages and some comments.
Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Move RCU Kconfig options from top-level menu to an "RCU Subsystem"
menu under the "General Setup" menu.
Signed-off-by: Mike Travis <travis@sgi.com>
Tested-by: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Andrew Morton reported this warning on ia64:
kernel/sched.c: In function `sd_init_NODE':
kernel/sched.c:7449: warning: comparison of distinct pointer types lacks a cast
Using the untyped min() function produces such warnings.
Fix: type the constant 32 as unsigned int to match typeof(num_online_cpus).
Reported-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: fix build warning
The macro cpu_to_node did not reference it's argument, and instead
simply returned a 0. This causes a "unused variable" warning if
it's the only reference in a function (show_cache_disable).
Replace it with the more correct inline function.
Signed-off-by: Mike Travis <travis@sgi.com>
Impact: fix build errors
Since the SPARSE IRQS changes redefined how the kstat irqs are
organized, arch's must use the new accessor function:
kstat_incr_irqs_this_cpu(irq, DESC);
If CONFIG_SPARSE_IRQS is set, then DESC is a pointer to the
irq_desc which has a pointer to the kstat_irqs. If not, then
the .irqs field of struct kernel_stat is used instead.
Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: cleanup, update to new cpumask API
Irq_desc.affinity and irq_desc.pending_mask are now cpumask_var_t's
so access to them should be using the new cpumask API.
Signed-off-by: Mike Travis <travis@sgi.com>
Ingo Molnar wrote:
> All non-x86 architectures fail to build:
>
> In file included from /home/mingo/tip/include/linux/random.h:11,
> from /home/mingo/tip/include/linux/stackprotector.h:6,
> from /home/mingo/tip/init/main.c:17:
> /home/mingo/tip/include/linux/irqnr.h:26:63: error: asm/irq_vectors.h: No such file or directory
Do not include asm/irq_vectors.h in generic code - it's not available
on all architectures.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: reduce memory usage.
Allocate kstat_irqs_legacy based on nr_cpu_ids to deal with this
memory usage bump when NR_CPUS bumped from 128 to 4096:
8192 +253952 262144 +3100% kstat_irqs_legacy(.bss)
This is only when CONFIG_SPARSE_IRQS=y.
Signed-off-by: Mike Travis <travis@sgi.com>
Impact: Reduce memory usage.
This is the second half of the changes to make the irq_desc_ptrs be
variable sized based on nr_cpu_ids. This is done by adding a new
"max_nr_irqs" macro to irq_vectors.h (and a dummy in irqnr.h) to
return a max NR_IRQS value based on NR_CPUS or nr_cpu_ids.
This necessitated moving the define of MAX_IO_APICS to a separate
file (asm/apicnum.h) so it could be included without the baggage
of the other asm/apicdef.h declarations.
Signed-off-by: Mike Travis <travis@sgi.com>
Impact: allocate irq_desc_ptrs in preparation for making it variable-sized.
This addresses this memory usage bump when NR_CPUS bumped from 128 to 4096:
34816 +229376 264192 +658% irq_desc_ptrs(.data.read_mostly)
The patch is split into two parts, the first simply allocates the
irq_desc_ptrs array. Then next will deal with making it variable.
This is only when CONFIG_SPARSE_IRQS=y.
Signed-off-by: Mike Travis <travis@sgi.com>
Impact: cleanup WARN msg.
Ingo requested:
> While at it, could you please also convert this to a WARN() construct
> instead? (in a separate commit)
... and it shall be done. ;-)
Signed-off-by: Mike Travis <travis@sgi.com>
Impact: preparation, cleanup, add KERN_INFO printk
Modify references from NR_IRQS to nr_irqs as the later will become
variable-sized based on nr_cpu_ids when CONFIG_SPARSE_IRQS=y.
Signed-off-by: Mike Travis <travis@sgi.com>
Impact: reduce stack usage, cleanup
Use a cpumask_var_t in find_lowest_rq() and clean up other old
cpumask_t calls.
Signed-off-by: Mike Travis <travis@sgi.com>
Impact: reduce stack usage.
init_intel_cacheinfo() does not use the cpumask so define a subset
of struct _cpuid4_info (_cpuid4_info_regs) that can be used instead.
Signed-off-by: Mike Travis <travis@sgi.com>
Impact: Reduce memory usage, use new cpumask API.
Use cpumask_var_t for 'cpus' cpumask in struct threshold_bank and update
remaining old cpumask_t functions to new cpumask API.
Signed-off-by: Mike Travis <travis@sgi.com>
Impact: Improve tlb flush performance for UV
Calling alloc_cpumask_var a zillion times a second does affect
performance. Replace with static cpumask.
Note: when CONFIG_X86_UV is defined, this extra PER_CPU memory
will be optimized out for non-UV configs as is_uv_system() will
then return a constant 0.
Signed-off-by: Mike Travis <travis@sgi.com>
Impact: fix bug where new irq_desc uses old cpumask pointers which are freed.
As Yinghai pointed out, init_copy_one_irq_desc() copies the old desc to
the new desc overwriting the cpumask pointers. Since the old_desc and
the cpumask pointers are freed, then memory corruption will occur if
these old pointers are used.
Move the allocation of these pointers to after the copy.
Signed-off-by: Mike Travis <travis@sgi.com>
Cc: Yinghai Lu <yinghai@kernel.org>
Impact: reduce memory usage, use new cpumask API.
Replace the affinity and pending_masks with cpumask_var_t's. This adds
to the significant size reduction done with the SPARSE_IRQS changes.
The added functions (init_alloc_desc_masks & init_copy_desc_masks) are
in the include file so they can be inlined (and optimized out for the
!CONFIG_CPUMASKS_OFFSTACK case.) [Naming chosen to be consistent with
the other init*irq functions, as well as the backwards arg declaration
of "from, to" instead of the more common "to, from" standard.]
Includes a slight change to the declaration of struct irq_desc to embed
the pending_mask within ifdef(CONFIG_SMP) to be consistent with other
references, and some small changes to Xen.
Tested: sparse/non-sparse/cpumask_offstack/non-cpumask_offstack/nonuma/nosmp on x86_64
Signed-off-by: Mike Travis <travis@sgi.com>
Cc: Chris Wright <chrisw@sous-sol.org>
Cc: Jeremy Fitzhardinge <jeremy@xensource.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Cc: virtualization@lists.osdl.org
Cc: xen-devel@lists.xensource.com
Cc: Yinghai Lu <yhlu.kernel@gmail.com>
Fix the following build errors reported by Yinghai Lu:
| In file included from arch/x86/mach-generic/summit.c:16:
| tip/linux-2.6/arch/x86/include/asm/summit/apic.h:
| In function 'cpu_mask_to_apicid_and':
| tip/linux-2.6/arch/x86/include/asm/summit/apic.h:179:
| error: 'GFP_ATOMIC' undeclared (first use in this function)
Reported-by: Yinghai Lu <Yinghai.Lu@Sun.COM>
Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: use new cpumask API to reduce stack usage
Replace the saving of current->cpus_allowed and set_cpus_allowed_ptr() with
a work_on_cpu function for read_measured_perf_ctrs().
Basically splits off the work function from get_measured_perf which is
run on the designated cpu. Moves definition of struct perf_cur out of
function local namespace, and is used as the work function argument.
References in get_measured_perf use values in the perf_cur struct.
Signed-off-by: Mike Travis <travis@sgi.com>
Acked-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: use new cpumask API to reduce stack usage
Replace the saving of current->cpus_allowed and set_cpus_allowed_ptr() with
a work_on_cpu function for drv_read() and drv_write().
Basically converts do_drv_{read,write} into "work_on_cpu" functions that
are now called by drv_read and drv_write.
Signed-off-by: Mike Travis <travis@sgi.com>
Acked-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: cleanup, reduce stack usage, use new cpumask API.
Replace the cpumask_t in struct drv_cmd with a cpumask_var_t. Remove unneeded
online_policy_cpus cpumask_t in acpi_cpufreq_target. Update refs to use
new cpumask API.
Signed-off-by: Mike Travis <travis@sgi.com>
Acked-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: use new cpumask API to reduce stack usage
Replace the saving of current->cpus_allowed and set_cpus_allowed_ptr() with
a work_on_cpu function for the acpi_processor_ffh_cstate_probe() function.
Basically splits acpi_processor_ffh_cstate_probe() into two functions, the
other being acpi_processor_ffh_cstate_probe_cpu which is the work function
run on the designated cpu.
Signed-off-by: Mike Travis <travis@sgi.com>
Acked-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: use new cpumask API to reduce memory and stack usage
Allocate the following local cpumasks based on the number of cpus that
are present. References will use new cpumask API. (Currently only
modified for x86_64, x86_32 continues to use the *_map variants.)
cpu_callin_mask
cpu_callout_mask
cpu_initialized_mask
cpu_sibling_setup_mask
Provide the following accessor functions:
struct cpumask *cpu_sibling_mask(int cpu)
struct cpumask *cpu_core_mask(int cpu)
Other changes are when setting or clearing the cpu online, possible
or present maps, use the accessor functions.
Signed-off-by: Mike Travis <travis@sgi.com>
Acked-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: use new cpumask API to reduce stack usage
Replace the local cpumask_t variable with a pointer to the
const cpumask that needs to be printed.
Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: use new cpumask API to reduce stack usage
Replace the local cpumask_t variable with a pointer to the
const cpumask that needs to be printed.
Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: cleanup
The function prototype should use 'struct cpumask *' to declare
cpumask arguments (instead of cpumask_var_t).
Note: arch/ia64/kernel/irq.c still had the following "old cpumask_t" usages:
105: cpumask_t mask = CPU_MASK_NONE;
107: cpu_set(cpu_logical_id(hwid), mask);
110: irq_desc[irq].affinity = mask;
... replaced with a simple "cpumask_of(cpu_logical_id(hwid))".
161: new_cpu = any_online_cpu(cpu_online_map);
194: time_keeper_id = first_cpu(cpu_online_map);
... replaced with cpu_online_mask refs.
Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
CONFIG_NR_CPUS will be defined for all arch's whether SMP or not, but
it may not have made it into all arches yet.
Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: Reduce inter-node memory traffic.
Reduces inter-node memory traffic (offloading the global system bus)
by allocating referenced struct cpumasks on the same node as the
referring struct.
Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: Reduce future system panics due to cpumask operations using NR_CPUS
Insure that code does not look at bits >= nr_cpu_ids as when cpumasks are
allocated based on nr_cpu_ids, these extra bits will not be defined.
Also some other minor updates:
* change in to use cpu accessor function set_cpu_present() instead of
directly accessing cpu_present_map w/cpu_clear() [arch/x86/kernel/reboot.c]
* use cpumask_of() instead of &cpumask_of_cpu() [arch/x86/kernel/reboot.c]
* optimize some cpu_mask_to_apicid_and functions.
Signed-off-by: Mike Travis <travis@sgi.com>
Acked-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: prevents panic from stack overflow on numa-capable machines.
Some of the "removal of stack hogs" changes in kernel/sched.c by using
node_to_cpumask_ptr were undone by the early cpumask API updates, and
causes a panic due to stack overflow. This patch undoes those changes
by using cpumask_of_node() which returns a 'const struct cpumask *'.
In addition, cpu_coregoup_map is replaced with cpu_coregroup_mask further
reducing stack usage. (Both of these updates removed 9 FIXME's!)
Also:
Pick up some remaining changes from the old 'cpumask_t' functions to
the new 'struct cpumask *' functions.
Optimize memory traffic by allocating each percpu local_cpu_mask on the
same node as the referring cpu.
Signed-off-by: Mike Travis <travis@sgi.com>
Acked-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: enables /sys/devices/system/cpu/{kernel_max,offline} user interface
By setting total_cpus, the drivers/base/cpu.c will display the
values of kernel_max (NR_CPUS-1) and the offlined cpu map.
Signed-off-by: Mike Travis <travis@sgi.com>
Acked-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Add information to cputopology.txt explaining the output of various
system cpumask's.
Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Acked-by: Greg Kroah-Hartman <gregkh@suse.de>
Impact: add new sysfs files.
Add sysfs files "kernel_max" and "offline" to display the max CPU index
allowed (NR_CPUS-1), and the map of cpus that are offline.
Cpus can be offlined via HOTPLUG, disabled by the BIOS ACPI tables, or
if they exceed the number of cpus allowed by the NR_CPUS config option,
or the "maxcpus=NUM" kernel start parameter.
The "possible_cpus=NUM" parameter can also extend the number of possible
cpus allowed, in which case the cpus not present at startup will be
in the offline state. (These cpus can be HOTPLUGGED ON after system
startup [pending a follow-on patch to provide the capability via the
/sys/devices/sys/cpu/cpuN/online mechanism to bring them online.])
By design, the "offlined cpus > possible cpus" display will always
use the following formats:
* all possible cpus online: "x$" or "x-y$"
* some possible cpus offline: ".*,x$" or ".*,x-y$"
where:
x == number of possible cpus (nr_cpu_ids); and
y == number of cpus >= NR_CPUS or maxcpus (if y > x).
One use of this feature is for distros to select (or configure) the
appropriate kernel to install for the resident system.
Notes:
* cpus offlined <= possible cpus will be printed for all architectures.
* cpus offlined > possible cpus will only be printed for arches that
set 'total_cpus' [X86 only in this patch].
Based on tip/cpus4096 + .../rusty/linux-2.6-for-ingo.git/master +
x86-only-patches sent 12/15.
Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Impact: New kerneldoc comments
Additional documentation added to all the alloc_cpumask and free_cpumask
functions.
Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> (minor additions)
Impact: New API
This will be needed in x86 code to allocate the domain and old_domain
cpumasks on the same node as where the containing irq_cfg struct is
allocated.
(Also fixes double-dump_stack on rare CONFIG_DEBUG_PER_CPU_MAPS case)
Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> (re-impl alloc_cpumask_var)
Impact: add new boot parameter
Use possible_cpus=NUM kernel parameter to extend the number of possible
cpus.
The ability to HOTPLUG ON cpus that are "possible" but not "present" is
dealt with in a later patch.
Signed-off-by: Mike Travis <travis@sgi.com>
Impact: fix potential APIC crash
In determining the destination apicid, there are usually three cpumasks
that are considered: the incoming cpumask arg, cfg->domain and the
cpu_online_mask. Since we are just introducing the cpu_mask_to_apicid_and
function, make sure it includes the cpu_online_mask in it's evaluation.
[Added with this patch.]
There are two io_apic.c functions that did not previously use the
cpu_online_mask: setup_IO_APIC_irq and msi_compose_msg. Both of these
simply used cpu_mask_to_apicid(cfg->domain & TARGET_CPUS), and all but
one arch (NUMAQ[*]) returns only online cpus in the TARGET_CPUS mask,
so the behavior is identical for all cases.
[*: NUMAQ bug?]
Note that alloc_cpumask_var is only used for the 32-bit cases where
it's highly likely that the cpumask set size will be small and therefore
CPUMASK_OFFSTACK=n. But if that's not the case, failing the allocate
will cause the same return value as the default.
Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: new API
The old topology_core_siblings() and topology_thread_siblings() return
a cpumask_t; these new ones return a (const) struct cpumask *.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Mike Travis <travis@sgi.com>
Impact: use new API, remove cpumask from stack.
Change smp_call_function_mask() callers to smp_call_function_many().
This removes a cpumask from the stack, and falls back should allocating
the cpumask var fail (only possible with CONFIG_CPUMASKS_OFFSTACK).
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Mike Travis <travis@sgi.com>
Cc: jeremy@xensource.com
Impact: remove cpumask_t from stack.
We should not try to save and restore cpus_allowed on current.
We can't use work_on_cpu() here, since it's in the hotplug cpu path
(if anyone else tries to get the hotplug lock from a workqueue we
could deadlock against them).
Fortunately, we can just use smp_call_function_single() since the
function can run from an interrupt.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Mike Travis <travis@sgi.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Oleg Nesterov <oleg@tv-sign.ru>
Impact: use new API
Use the accessors rather than frobbing bits directly. Most of this is
in arch code I haven't even compiled, but is straightforward.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Mike Travis <travis@sgi.com>
Impact: cleanup, futureproof
In fact, all cpumask ops will only be valid (in general) for bit
numbers < nr_cpu_ids. So use that instead of NR_CPUS in various
places.
This is always safe: no cpu number can be >= nr_cpu_ids, and
nr_cpu_ids is initialized to NR_CPUS at boot.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Mike Travis <travis@sgi.com>
Acked-by: Ingo Molnar <mingo@elte.hu>
This patch simply changes cpumask_t to struct cpumask and similar
trivial modernizations.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Mike Travis <travis@sgi.com>
Impact: cleanup, remove on-stack cpumask.
The "map" arg is always cpu_online_mask. Importantly, set_affinity
always ands the argument with cpu_online_mask anyway, so we don't need
to do it in fixup_irqs(), avoiding a temporary.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Mike Travis <travis@sgi.com>
Impact: cleanup, consolidate patches, use new API
Consolidate the following into a single patch to adapt to new
sparseirq code in arch/x86/kernel/io_apic.c, add allocation of
cpumask_var_t's in domain and old_domain, and reduce further
merge conflicts. Only one file (arch/x86/kernel/io_apic.c) is
changed in all of these patches.
0006-x86-io_apic-change-irq_cfg-domain-old_domain-to.patch
0007-x86-io_apic-set_desc_affinity.patch
0008-x86-io_apic-send_cleanup_vector.patch
0009-x86-io_apic-eliminate-remaining-cpumask_ts-from-st.patch
0021-x86-final-cleanups-in-io_apic-to-use-new-cpumask-AP.patch
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Mike Travis <travis@sgi.com>
Mike Travis