The function graph tracer is currently the most invasive tracer
in the ftrace family. It can easily overflow the buffer even with
10megs per CPU. This means that events can often be lost.
On start up, or after events are lost, if the function return is
recorded but the function enter was lost, all we get to see is the
exiting '}'.
Here is how a typical trace output starts:
[tracing] cat trace
# tracer: function_graph
#
# CPU DURATION FUNCTION CALLS
# | | | | | | |
0) + 91.897 us | }
0) ! 567.961 us | }
0) <========== |
0) ! 579.083 us | _raw_spin_lock_irqsave();
0) 4.694 us | _raw_spin_unlock_irqrestore();
0) ! 594.862 us | }
0) ! 603.361 us | }
0) ! 613.574 us | }
0) ! 623.554 us | }
0) 3.653 us | fget_light();
0) | sock_poll() {
There are a series of '}' with no matching "func() {". There's no information
to what functions these ending brackets belong to.
This patch adds a stack on the per cpu structure used in outputting
the function graph tracer to keep track of what function was outputted.
Then on a function exit event, it checks the depth to see if the
function exit has a matching entry event. If it does, then it only
prints the '}', otherwise it adds the function name after the '}'.
This allows function exit events to show what function they belong to
at trace output startup, when the entry was lost due to ring buffer
overflow, or even after a new task is scheduled in.
Here is what the above trace will look like after this patch:
[tracing] cat trace
# tracer: function_graph
#
# CPU DURATION FUNCTION CALLS
# | | | | | | |
0) + 91.897 us | } (irq_exit)
0) ! 567.961 us | } (smp_apic_timer_interrupt)
0) <========== |
0) ! 579.083 us | _raw_spin_lock_irqsave();
0) 4.694 us | _raw_spin_unlock_irqrestore();
0) ! 594.862 us | } (add_wait_queue)
0) ! 603.361 us | } (__pollwait)
0) ! 613.574 us | } (tcp_poll)
0) ! 623.554 us | } (sock_poll)
0) 3.653 us | fget_light();
0) | sock_poll() {
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
With this patch the prefix registers of all online CPUs are stored in the
the zcore dump header. This allows dump analysis tools to access the register
information that is stored in the prefix pages without using the System.map.
Signed-off-by: Michael Holzheu <holzheu@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
The glibc vdso code for s390 uses the version string 2.6.29, the
kernel uses the version string 2.6.26. No wonder the vdso code
is never used. The first kernel version to contain the vdso code
is 2.6.29 which makes this the correct version.
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Add the "bzImage" compile target and the necessary code to generate
compressed kernel images. The old style uncompressed "image" target
is preserved, a simple make will build them both.
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
S390 ELF core dump currently only contains the PSW, the general purpose
registers, the floating point registers and the access registers stored
in PRSTATUS/PRFPREG note sections.
For analyzing s390 kernel problems additional registers are important.
In order to be able to include these registers to a kernel ELF core dump,
this patch adds the following five new note sections to elf.h:
* NT_S390_TIMER: S390 timer register
* NT_S390_TODCMP: S390 TOD comparator register
* NT_S390_TODPREG: S390 TOD programmable register
* NT_S390_CTRS: S390 control registers
* NT_S390_PREFIX: S390 prefix register
The new note sections have been already defined and accepted in the upstream
binutils package.
Signed-off-by: Michael Holzheu <holzheu@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Move the ebcdic to ascii conversion of the kernel parameter line from
head.S to early.c and convert the assembler code to C.
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Add the instruction of the z9-ec and z10 machines to the kernel disassembler.
Add the missing "ptff" instruction of z9-109 and the missing "sqd" of g5.
Remove useless comments with instruction examples from format table.
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Flushing the dasd ccw request queue may stop the processing of the
block device request queue. Destroy partitions may wait for
outstanding requests and thus hang.
Swapping dasd_destroy_partitions and dasd_flush_request_queue so that
the request queue is empty before dasd_destroy_partitions is called.
Signed-off-by: Stefan Haberland <stefan.haberland@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
The function dasd_device_from_cdev returns a reference to the dasd
device and increases the refcount by one. If an exception occurs,
the refcount was not decreased in all cases
e.g. in dasd_discipline_show.
Prevent the offline processing from hang by correcting two functions
to decrease the refcount even if an error occured.
Signed-off-by: Stefan Haberland <stefan.haberland@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Setting a DASD online and offline in quick succession may cause
a kernel panic or let the chhccwdev command wait forever.
The Online process is split into two parts. After the first part
is finished the offline process may be called. This may result
in a situation where the second online processing part tries to
set the DASD offline as well.
Use a mutex to protect online and offline against each other.
Also correct some checking.
Signed-off-by: Stefan Haberland <stefan.haberland@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Use kprobes_built_in() to avoid ifdefs like most other architectures do.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Reduces the size of the bug table entries by 50% on 64bit kernels.
Saves around 30kb on a defconfig kernel.
s390 version of b93a531e "allow bug table entries to use relative
pointers (and use it on x86-64)".
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Use asm offsets to make sure the offset defines to struct _lowcore and
its layout don't get out of sync.
Also add a BUILD_BUG_ON() which checks that the size of the structure
is sane.
And while being at it change those sites which use odd casts to access
the current lowcore. These should use S390_lowcore instead.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
free_initmem() and free_initrd_mem() are nearly identical. So make them
call a common function.
Also fixes a bug: if the initrd wouldn't start on a page boundary also
memory after the initrd would be initialized with the poison value.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
ENOTSUPP is not supposed to leak to userspace so lets just use
EOPNOTSUPP everywhere.
Doesn't fix a bug, but makes future reviews easier.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
This patch introduces a new function that checks the running status
of a cpu in a hypervisor. This status is not virtualized, so the check
is only correct if running in an LPAR. On acquiring a spinlock, if the
cpu holding the lock is scheduled by the hypervisor, we do a busy wait
on the lock. If it is not scheduled, we yield over to that cpu.
Signed-off-by: Gerald Schaefer <gerald.schaefer@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
The size of the field that contains the description block count is
only four bits instead of eight bits.
The first four bits are reserved but this might change and break.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Introduce the MACHINE_IS_LPAR flag for code that should only be
executed if Linux is running in an LPAR.
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Remove a memset hack that relied on the internal layout of the
qdio_irq struct and move the per device statistics data into an own
cache line to avoid cache line bashing between the inbound and the
outbound queue tasklets. Also reduce the number of allocated queues
from 32 to 4 which is the current maximum. That saves a cache line
in struct qdio_irq.
Signed-off-by: Jan Glauber <jang@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Add counters for the number of processed SBALs. The numbers summarize
how many SBALs were processed at each queue scan and indicate the
utilization of the queue. Furthermore the number of unsuccessfull
queue scans, SBAL errors and the total number of processed
SBALs are accounted.
Also regroup struct qdio_q to move read-mostly and write-mostly data
into different cachelines.
Signed-off-by: Jan Glauber <jang@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Rename signal_processor* functions to sigp*.
Add raw variants of each version, so we can get rid of the hacks played
in smp code which establish temporary cpu logical mappings so they could
call the sigp functions.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Always reboot on logical cpu 0. This makes sure that the IPL cpu is
always the same and usually avoids strange numbering schemes between
physical and logical cpus.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Rename __LC_RESTART_PSW to __LC_RST_NEW_PSW, add a define for the
missing 32 bit variant and the missing old PSWs.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Remove support to be able to dump 31 bit systems with a 64 bit dumper.
This is mostly useless since no distro ships 31 bit kernels together
with a 64 bit dumper.
We also get rid of a bit of hacky code.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Drop support to compile the kernel with gcc versions older than 3.3.3.
This allows us to use the "Q" inline assembly contraint on some more
inline assemblies without duplicating a lot of complex code (e.g. __xchg
and __cmpxchg). The distinction for older gcc versions can be removed
which saves a few lines and simplifies the code.
Reviewed-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Some parts of cio do not shift PAGE_DEFAULT_KEY correctly and end up
with an incorrect key in their data structures.
Since the default key is zero this doesn't really matter. However if
somebody would use key-controlled protection for debugging purposes
it would be quite helpful if all of this would work as expected.
Also remove a stale declaration.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
ccw_device_pm_restore: trigger subchannel event to better handle
changes to the subchannel device.
Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Callers of ccw_device_notify could not distinguish between a driver
who has no notifier registered and a driver who doesn't want to keep
a device after a certain event. Change this by adding proper return
codes.
Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Make the potentially long blocking wait_event's used by the cio
settle mechanism interruptible.
Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
To fetch a pending channel report word (crw) we use a kernel
thread which triggers stcrw and sleeps on a semaphore. The s390
machine check handler uses crw_handle_channel_report to handle
one crw if needed.
This patch replaces the semaphore with a waitqueue (to block the
kernel thread) and an atomic_t (to count the number of pending
requests).
By this we achieve the ability to force this thread to check for
a pending crw (independent on when it is triggered by the machine
check handler) and wait for this action to finish.
Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
This patch introduces a proc file cio_settle. A write request to
this file is blocked until all queued cio actions are handled.
This will allow userspace to wait for pending work affecting
device availability after changing cio_ignore or the hardware
configuration.
Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
We used to maintain 2 singlethreaded workqueues for synchronization
and to trigger work from interrupt context. Since our latest cio
changes we only use one of these workqueues. So get rid of the
unused workqueue, rename the remaining one to "cio_work_q" and move
its ownership to the channel subsystem driver.
Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
cu3088 layer for lcs and ctcm has been removed. Thus the reference
to cu3088 in this text is to be removed as well.
Signed-off-by: Ursula Braun <ursula.braun@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Same as on x86 and sparc, besides the fact that enabling the option
will just emit compile time warnings instead of errors.
Keeps allyesconfig kernels compiling.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Convert netmask to __be32 and format it with %pI4
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Konrad Rzeszutek <ketuzsezr@darnok.org>
Reviewed-by: Mike Christie <michaelc@cs.wisc.edu>
Cc: Peter Jones <pjones@redhat.com>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
The hibernate memory preallocation code allocates memory to push some
user space data out of physical RAM, so that the hibernation image is
not too large. It allocates more memory than necessary for creating
the image, so it has to release some pages to make room for
allocations made while suspending devices and disabling nonboot CPUs,
or the system will hang due to the lack of free pages to allocate
from. Unfortunately, the function used for freeing these pages,
free_unnecessary_pages(), contains a bug that prevents it from doing
the job on all systems without highmem.
Fix this problem, which is a regression from the 2.6.30 kernel, by
using the right condition for the termination of the loop in
free_unnecessary_pages().
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Reported-and-tested-by: Alan Jenkins <sourcejedi.lkml@googlemail.com>
Cc: stable@kernel.org
Its contents and entry in Makefile were already removed in
8e60c6a134
(Shift remaining code from swsusp.c to hibernate.c)
but somehow it remained in-place (rjw: which most likely was my
mistake).
Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Acked-by: Nigel Cunningham <nigel@tuxonice.net>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Remove a trailing space from a message in swsusp_save().
Signed-off-by: Frans Pop <elendil@planet.nl>
Acked-by: Pavel Machek <pavel@ucw.cz>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Set power.async_suspend for all SCSI devices, targets and hosts, so
that they can be suspended and resumed in parallel with the main
suspend/resume thread and possibly with other devices they don't
depend on in a known way (i.e. devices which are not their parents or
children).
The power.async_suspend flag is also set for devices that don't have
suspend or resume callbacks, because otherwise they would make the
main suspend/resume thread wait for their "asynchronous" children
(during suspend) or parents (during resume), effectively negating the
possible gains from executing these devices' suspend and resume
callbacks asynchronously.
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Set power.async_suspend for USB devices, endpoints and interfaces,
allowing them to be suspended and resumed asynchronously during
system sleep transitions.
The power.async_suspend flag is also set for devices that don't have
suspend or resume callbacks, because otherwise they would make the
main suspend/resume thread wait for their "asynchronous" children
(during suspend) or parents (during resume), effectively negating the
possible gains from executing these devices' suspend and resume
callbacks asynchronously.
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
This patch (as1331) adds non-tree ordering constraints needed for
proper resume of PCI USB host controllers from hibernation. The main
issue is that non-high-speed devices must not be resumed before the
high-speed root hub, because it is the ehci_bus_resume() routine which
takes care of handing the device connection over to the companion
controller. If the device resume is attempted before the handover
then the device won't be found and it will be treated as though it had
disconnected.
The patch adds a new field to the usb_bus structure; for each
full/low-speed bus this field will contain a pointer to the companion
high-speed bus (if one exists). It is used during normal device
resume; if the hs_companion pointer isn't NULL then we wait for the
root-hub device on the hs_companion bus.
A secondary issue is that an EHCI controlller shouldn't be resumed
before any of its companions. On some machines I have observed
handovers failing if the companion controller is reinitialized after
the handover. Thus, the EHCI resume routine must wait for the
companion controllers to be resumed.
The patch also fixes a small bug in usb_hcd_pci_probe(); an error path
jumps to the wrong label, causing a memory leak.
[rjw: Fixed compilation for CONFIG_PM_SLEEP unset.]
Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Acked-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Set power.async_suspend for all PCI devices and PCIe port services,
so that they can be suspended and resumed in parallel with other
devices they don't depend on in a known way (i.e. devices which are
not their parents or children).
This only affects the "regular" suspend and resume stages, which
means in particular that the restoration of the PCI devices' standard
configuration registers during resume will still be carried out
synchronously (at the "early" resume stage).
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
It will never reach here if the sws_resume_bdev is erratic.
swsusp_read() is called only from software_resume(), but after
swsusp_check() which would catch the error state.
Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
They were deprecated and removed from exported headers more than 2
years ago. Inform users about their removal in the future now.
(Switch cases needed to be reorderded for an easy fall through.)
And add an entry to feature-removal-schedule.
Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Acked-by: Pavel Machek <pavel@ucw.cz>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
There are some dependencies between devices (in particular, between
EHCI USB controllers and their OHCI/UHCI siblings) which are not
reflected by the structure of the device tree. With synchronous
suspend and resume these dependencies are taken into accout
automatically, because the devices in question are always registered
in the right order, but to meet these constraints with asynchronous
suspend and resume the drivers of these devices will need to use
dpm_wait() in their suspend/resume routines, so introduce a helper
function allowing them to do that.
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
