Ingo Molnar [Mon, 22 Jun 2015 12:40:03 +0000 (14:40 +0200)]
x86: Load __USER_DS into DS/ES after resume
Srinivas Pandruvada reported a problem with system resume from
suspend-to-RAM on 32-bit x86 systems where the DS register of
the CPU is set to __KERNEL_DS instead of __USER_DS on return
to user space which cases a General Protection Fault to occur.
The issue is that DS is set to __KERNEL_DS by the ACPI resume code
path while the SYSEXIT path never reloads DS/ES. It assumes they
are still __USER_DS set at the SYSENTER time (Brian Gerst), so if
the return to user space happens to be through SYSEXIT, it will lead
to the reported GPF.
Fix the problem by setting the DS and ES registers to __USER_DS
as expected by the SYSEXIT path.
* pm-opp:
PM / OPP: Add binding for 'opp-suspend'
PM / OPP: Allow multiple OPP tables to be passed via DT
PM / OPP: Add new bindings to address shortcomings of existing bindings
Viresh Kumar [Mon, 22 Jun 2015 12:21:38 +0000 (14:21 +0200)]
PM / OPP: Allow multiple OPP tables to be passed via DT
On some platforms (Like Qualcomm's SoCs), it is not decided until
runtime on what OPPs to use. The OPP tables can be fixed at compile
time, but which table to use is found out only after reading some efuses
(sort of an prom) and knowing characteristics of the SoC.
To support such platform we need to pass multiple OPP tables per device
and hardware should be able to choose one and only one table out of
those.
Update operating-points-v2 bindings to support that.
Viresh Kumar [Thu, 4 Jun 2015 16:20:31 +0000 (21:50 +0530)]
PM / OPP: Add new bindings to address shortcomings of existing bindings
Current OPP (Operating performance point) device tree bindings have been
insufficient due to the inflexible nature of the original bindings. Over
time, we have realized that Operating Performance Point definitions and
usage is varied depending on the SoC and a "single size (just frequency,
voltage) fits all" model which the original bindings attempted and
failed.
The proposed next generation of the bindings addresses by providing a
expandable binding for OPPs and introduces the following common
shortcomings seen with the original bindings:
- Getting clock/voltage/current rails sharing information between CPUs.
Shared by all cores vs independent clock per core vs shared clock per
cluster.
- Support for specifying current levels along with voltages.
- Support for multiple regulators.
- Support for turbo modes.
- Other per OPP settings: transition latencies, disabled status, etc.?
- Expandability of OPPs in future.
This patch introduces new bindings "operating-points-v2" to get these problems
solved. Refer to the bindings for more details.
We now have multiple versions of OPP binding and only one of them should
be used per device.
Mika Westerberg [Tue, 9 Jun 2015 09:17:07 +0000 (12:17 +0300)]
mfd: lpc_ich: Assign subdevice ids automatically
Using -1 as platform device id means that the platform driver core will not
assign any id to the device (the device name will not have id at all). This
results problems on systems that have multiple PCHs (Platform Controller
HUBs) because all of them also include their own copy of LPC device.
All the subsequent device creations will fail because there already exists
platform device with the same name.
Fix this by passing PLATFORM_DEVID_AUTO as platform device id. This makes
the platform device core to allocate new ids automatically.
Fabio Estevam [Sat, 16 May 2015 18:42:17 +0000 (15:42 -0300)]
mfd: si476x-i2c: Pass the IRQF_ONESHOT flag
Since commit 1c6c69525b40eb76de8adf039409722015927dc3 ("genirq: Reject
bogus threaded irq requests") threaded IRQs without a primary handler
need to be requested with IRQF_ONESHOT, otherwise the request will fail.
So pass the IRQF_ONESHOT flag in this case.
The semantic patch that makes this change is available
in scripts/coccinelle/misc/irqf_oneshot.cocci.
Fabio Estevam [Sat, 16 May 2015 18:42:16 +0000 (15:42 -0300)]
mfd: ab8500-gpadc: Pass the IRQF_ONESHOT flag
Since commit 1c6c69525b40eb76de8adf039409722015927dc3 ("genirq: Reject
bogus threaded irq requests") threaded IRQs without a primary handler
need to be requested with IRQF_ONESHOT, otherwise the request will fail.
So pass the IRQF_ONESHOT flag in this case.
The semantic patch that makes this change is available
in scripts/coccinelle/misc/irqf_oneshot.cocci.
Fabio Estevam [Sat, 16 May 2015 18:42:15 +0000 (15:42 -0300)]
mfd: ab8500-debugfs: Pass the IRQF_ONESHOT flag
Since commit 1c6c69525b40eb76de8adf039409722015927dc3 ("genirq: Reject
bogus threaded irq requests") threaded IRQs without a primary handler
need to be requested with IRQF_ONESHOT, otherwise the request will fail.
So pass the IRQF_ONESHOT flag in this case.
The semantic patch that makes this change is available
in scripts/coccinelle/misc/irqf_oneshot.cocci.
Fabio Estevam [Sat, 16 May 2015 18:42:14 +0000 (15:42 -0300)]
mfd: wm8350-core: Pass the IRQF_ONESHOT flag
Since commit 1c6c69525b40eb76de8adf039409722015927dc3 ("genirq: Reject
bogus threaded irq requests") threaded IRQs without a primary handler
need to be requested with IRQF_ONESHOT, otherwise the request will fail.
So pass the IRQF_ONESHOT flag in this case.
The semantic patch that makes this change is available
in scripts/coccinelle/misc/irqf_oneshot.cocci.
Fabio Estevam [Sat, 16 May 2015 18:42:13 +0000 (15:42 -0300)]
mfd: htc-i2cpld: Pass the IRQF_ONESHOT flag
Since commit 1c6c69525b40eb76de8adf039409722015927dc3 ("genirq: Reject
bogus threaded irq requests") threaded IRQs without a primary handler
need to be requested with IRQF_ONESHOT, otherwise the request will fail.
So pass the IRQF_ONESHOT flag in this case.
The semantic patch that makes this change is available
in scripts/coccinelle/misc/irqf_oneshot.cocci.
Fabio Estevam [Sat, 16 May 2015 18:42:12 +0000 (15:42 -0300)]
mfd: wm831x-auxadc: Pass the IRQF_ONESHOT flag
Since commit 1c6c69525b40eb76de8adf039409722015927dc3 ("genirq: Reject
bogus threaded irq requests") threaded IRQs without a primary handler
need to be requested with IRQF_ONESHOT, otherwise the request will fail.
So pass the IRQF_ONESHOT flag in this case.
The semantic patch that makes this change is available
in scripts/coccinelle/misc/irqf_oneshot.cocci.
Fabio Estevam [Sat, 16 May 2015 18:42:11 +0000 (15:42 -0300)]
mfd: mc13xxx-core: Pass the IRQF_ONESHOT flag
Since commit 1c6c69525b40eb76de8adf039409722015927dc3 ("genirq: Reject
bogus threaded irq requests") threaded IRQs without a primary handler
need to be requested with IRQF_ONESHOT, otherwise the request will fail.
So pass the IRQF_ONESHOT flag in this case.
The semantic patch that makes this change is available
in scripts/coccinelle/misc/irqf_oneshot.cocci.
Fabio Estevam [Sat, 16 May 2015 18:42:10 +0000 (15:42 -0300)]
mfd: twl4030-irq: Pass the IRQF_ONESHOT flag
Since commit 1c6c69525b40eb76de8adf039409722015927dc3 ("genirq: Reject
bogus threaded irq requests") threaded IRQs without a primary handler
need to be requested with IRQF_ONESHOT, otherwise the request will fail.
So pass the IRQF_ONESHOT flag in this case.
The semantic patch that makes this change is available
in scripts/coccinelle/misc/irqf_oneshot.cocci.
Steve Twiss [Tue, 19 May 2015 10:32:45 +0000 (11:32 +0100)]
mfd: da9063: Add support for OnKey driver
Add MFD support for the DA9063 OnKey driver
The function da9063_clear_fault_log() is added to mitigate the case of a
hardware power-cut after a long-long OnKey press. Although there is no
software intervention in this case (by definition) such a shutdown would
cause persistent information within the DA9063 FAULT_LOG that would be
available during the next device restart.
Clearance of this persistent register must be completed after such a
hardware power-cut operation has happened so that the FAULT_LOG does not
continue with previous values. The clearance function has been added here
in the kernel driver because wiping the fault-log cannot be counted on
outside the Linux kernel.
Signed-off-by: Steve Twiss <[email protected]>
[Lee: Removed 'key_power' for Dmitry to take through the Input Tree] Signed-off-by: Lee Jones <[email protected]>
The use of ifneq against 'n' to conditionally compile codec-specific
parts is wrong and was resulting in all the codec tables being built
even for deselected codecs.
For the WM5102 there is a dependency in the core code on wm5102_patch()
which only exists when CONFIG_MFD_WM5102 is defined. To avoid having
to sprinkle #ifdefs around the code it is given an alternative empty
stub version when CONFIG_MFD_WM5102 is deselected
mfd: Check ACPI device companion before checking resources
Current code in mfd-core calls into ACPI to check resources even
on a system that booted with a DT (on kernels with both DT and ACPI
support compiled in). This triggers ACPI exceptions since we may
end up calling the ACPI interpreter when it has not been initialized:
"ACPI Exception: AE_BAD_PARAMETER, Thread 2064154624 could not acquire
Mutex [0x1] (20150410/utmutex-285)"
This patch fixes the issues by adding a check for an ACPI companion
device before carrying out ACPI resources checks to avoid calling
the ACPI interpreter if the fwnode representing the device is an OF one.
Later arizona silicon has the single/differential selector
in a different register, and IN1_MODE only selects between
analogue or digital. Prepare for this by splitting the
INx_MODE definition into two fields.
Charles Keepax [Mon, 11 May 2015 12:58:09 +0000 (13:58 +0100)]
mfd: arizona: Add better support for system suspend
Allow the chip to completely power off if we enter runtime suspend and
there is no jack detection active. This is helpful for systems where
system suspend might remove the supplies to the CODEC, without informing
us. Note the powering off is done in runtime suspend rather than system
suspend, because we need to hold reset until the first time DCVDD is
powered anyway (which would be in runtime resume), and we might as well
save the extra power.
Charles Keepax [Mon, 11 May 2015 12:58:08 +0000 (13:58 +0100)]
mfd: wm5110: Set DCVDD voltage to 1.175V before entering sleep mode
The low power sleep mode on wm5110 requires that the LDO1 regulator be
set to 1.175V prior to entering sleep, then returned to 1.2V after
exiting sleep mode. This patch apply these regulator settings.
Charles Keepax [Mon, 11 May 2015 12:58:06 +0000 (13:58 +0100)]
mfd: wm5110: Add register patch required for low power sleep
Some register settings must be applied before the first time low power
sleep mode is entered on the wm5110 to ensure optimium performance.
These settings require SYSCLK to be enabled whilst they are being
applied. This patch applies the settings using the recently factored out
boot time SYSCLK functionality.
Charles Keepax [Mon, 11 May 2015 12:58:05 +0000 (13:58 +0100)]
mfd: arizona: Factor out SYSCLK enable from wm5102 hardware patch
wm5102 applies a custom hardware boot sequence, for this the SYSCLK
needs to be enabled. This patch factors out the code that enables
SYSCLK for this sequence such that it can be used for other boot time
operations that require SYSCLK.
Charles Keepax [Mon, 11 May 2015 12:58:03 +0000 (13:58 +0100)]
mfd: wm5102: Ensure we always boot the device fully
The wm5102 uses a custom boot sequence and the standard boot sequence is
disabled. However, the standard boot sequence must be run at least once
after a cold boot (caused by either a hard reset or removal of AVDD).
Unfortunately the register WRITE_SEQUENCER_CTRL_3 is not affected by the
hardware reset. This means if the device has been previously booted but
the AVDD supply has never been removed, arizona_dev_init will reset the
chip but the boot sequence will not run, which can cause numerous
problems. The solution is to manually clear this register and then boot
the chip again, which is what this patch does.
Lee Jones [Tue, 12 May 2015 13:13:58 +0000 (14:13 +0100)]
MAINTAINERS: Add Arnd Bergmann as Syscon reviewer
The System Configuration (syscon) driver is an important one, which
effects many ARM-SoC based platforms. I'm adding Arnd as a primary
reviewer as he is highly competent in this area and has a keen
interest in ARM-SoC.
MAINTAINERS: Extend Samsung MFD drivers entry and add Krzysztof Kozlowski
Extend the entry for Samsung MFD drivers for PMIC devices (Power
Management Integrated Circuit) with bindings documentation, clock
(clk-s2mps11.c) and RTC drivers (rtc-s5m.c).
These PMIC devices are used on many Exynos-based boards like Arndale
Octa (S2MPS11), Gear 2 (S2MPS14).
Add Krzysztof Kozlowski as a supporter for reviewing them.
I am not the author of these drivers. However I have recently
contributed to most of them and I have access to datasheets and hardware.
mfd: Allow compile test of GPIO consumers if !GPIOLIB
The GPIO subsystem provides dummy GPIO consumer functions if GPIOLIB is
not enabled. Hence drivers that depend on GPIOLIB, but use GPIO consumer
functionality only, can still be compiled if GPIOLIB is not enabled.
Relax the dependency on GPIOLIB if COMPILE_TEST is enabled, where
appropriate.
Tony Lindgren [Mon, 27 Apr 2015 17:18:14 +0000 (10:18 -0700)]
mfd: twl4030-power: Fix pmic for boards that need AC charger disabled
I noticed the PMIC configuration on 37xx-evm won't actually shut down
the voltages during off-idle. Turns out 37xx-evm needs the AC charger
state transitions disabled like we are doing for SDP and LDP in the
legacy booting case.
Let's fix this for device tree based booting by setting up the quirk
flag based on the compatible flag. And let's also use the existing
define for STARTON_CHG.
Note that SDP and EVM do not have the PMIC clken wired to gate the
the oscillator while LDP has.
Charles Keepax [Wed, 22 Apr 2015 13:34:48 +0000 (14:34 +0100)]
mfd: arizona: Remove redundant register sync
This soft reset used to be located after the register patch had been
applied, but has since moved to before the patch is applied. At the new
location there is no requirement to do a register sync as no register
writes will have happened yet.
Constify in various drivers configuration data which is not modified:
- regmap_irq_chip,
- individual regmap_irq's in array,
- regmap_config,
- irq_domain_ops,
Mark Brown [Mon, 22 Jun 2015 10:19:56 +0000 (11:19 +0100)]
Merge remote-tracking branches 'regulator/topic/of', 'regulator/topic/pwm', 'regulator/topic/qcom' and 'regulator/topic/soft-start' into regulator-next
Mark Brown [Mon, 22 Jun 2015 10:19:51 +0000 (11:19 +0100)]
Merge remote-tracking branches 'regulator/topic/88pm8607', 'regulator/topic/arizona', 'regulator/topic/cleanup' and 'regulator/topic/da9062' into regulator-next
The CESA driver calls phys_to_virt() which is not available on all
architectures.
Remove the depency on COMPILE_TEST to prevent building this driver on
non ARM architectures.
Herbert Xu [Mon, 22 Jun 2015 02:31:40 +0000 (10:31 +0800)]
crypto: algif_aead - Temporarily disable all AEAD algorithms
As the AEAD conversion is still ongoing, we do not yet wish to
export legacy AEAD implementations to user-space, as their calling
convention will change.
This patch actually disables all AEAD algorithms because some of
them (e.g., cryptd) will need to be modified to propagate this flag.
Subsequent patches will reenable them on an individual basis.
Herbert Xu [Mon, 22 Jun 2015 02:14:19 +0000 (10:14 +0800)]
crypto: af_alg - Forbid the use internal algorithms
The bit CRYPTO_ALG_INTERNAL was added to stop af_alg from accessing
internal algorithms. However, af_alg itself was never modified to
actually stop that bit from being used by the user. Therefore the
user could always override it by specifying the relevant bit in the
type and/or mask.
This patch silently discards the bit in both type and mask.
Herbert Xu [Sun, 21 Jun 2015 11:11:49 +0000 (19:11 +0800)]
crypto: seqiv - Add compatibility support without RNG
When seqiv is used in compatibility mode, this patch allows it
to function even when an RNG Is not available. It also changes
the RNG allocation for the new explicit seqiv interface so that
we only hold a reference to the RNG during initialisation.
Herbert Xu [Sun, 21 Jun 2015 11:11:47 +0000 (19:11 +0800)]
crypto: eseqiv - Offer normal cipher functionality without RNG
The RNG may not be available during early boot, e.g., the relevant
modules may not be included in the initramfs. As the RNG Is only
needed for IPsec, we should not let this prevent use of ciphers
without IV generators, e.g., for disk encryption.
This patch postpones the RNG allocation to the init function so
that one failure during early boot does not make the RNG unavailable
for all subsequent users of the same cipher.
More importantly, it lets the cipher live even if RNG allocation
fails. Of course we no longer offer IV generation and which will
fail with an error if invoked. But all other cipher capabilities
will function as usual.
Herbert Xu [Sun, 21 Jun 2015 11:11:46 +0000 (19:11 +0800)]
crypto: chainiv - Offer normal cipher functionality without RNG
The RNG may not be available during early boot, e.g., the relevant
modules may not be included in the initramfs. As the RNG Is only
needed for IPsec, we should not let this prevent use of ciphers
without IV generators, e.g., for disk encryption.
This patch postpones the RNG allocation to the init function so
that one failure during early boot does not make the RNG unavailable
for all subsequent users of the same cipher.
More importantly, it lets the cipher live even if RNG allocation
fails. Of course we no longer offer IV generation and which will
fail with an error if invoked. But all other cipher capabilities
will function as usual.
Herbert Xu [Sun, 21 Jun 2015 11:11:45 +0000 (19:11 +0800)]
crypto: user - Add CRYPTO_MSG_DELRNG
This patch adds a new crypto_user command that allows the admin to
delete the crypto system RNG. Note that this can only be done if
the RNG is currently not in use. The next time it is used a new
system RNG will be allocated.
Herbert Xu [Sun, 21 Jun 2015 11:11:43 +0000 (19:11 +0800)]
crypto: rng - Do not free default RNG when it becomes unused
Currently we free the default RNG when its use count hits zero.
This was OK when the IV generators would latch onto the RNG at
instance creation time and keep it until the instance is torn
down.
Now that IV generators only keep the RNG reference during init
time this scheme causes the default RNG to come and go at a high
frequencey. This is highly undesirable as we want to keep a single
RNG in use unless the admin wants it to be removed.
This patch changes the scheme so that the system RNG once allocated
is never removed unless a specifically requested.
Herbert Xu [Sun, 21 Jun 2015 11:11:41 +0000 (19:11 +0800)]
crypto: skcipher - Allow givencrypt to be NULL
Currently for skcipher IV generators they must provide givencrypt
as that is the whole point. We are currently replacing skcipher
IV generators with explicit IV generators. In order to maintain
backwards compatibility, we need to allow the IV generators to
still function as a normal skcipher when the RNG Is not present
(e.g., in the initramfs during boot). IOW everything but givencrypt
and givdecrypt will still work but those two will fail.
Therefore this patch assigns a default givencrypt that simply
returns an error should it be NULL.
Paul Gortmaker [Sat, 20 Jun 2015 23:02:32 +0000 (19:02 -0400)]
clocksource: Increase dependencies of timer-stm32 to limit build wreckage
This driver leaks out into arch/parisc builds that don't have
CONFIG_GENERIC_CLOCKEVENTS, leading to the following (truncated)
wreckage:
CC drivers/clocksource/timer-stm32.o
drivers/clocksource/timer-stm32.c:38:28: error: field 'evtdev' has incomplete type
drivers/clocksource/timer-stm32.c:44:19: warning: 'enum clock_event_mode' declared inside parameter list
drivers/clocksource/timer-stm32.c:44:19: warning: its scope is only this definition or declaration, which is probably not what you want
drivers/clocksource/timer-stm32.c:43:62: error: parameter 1 ('mode') has incomplete type
drivers/clocksource/timer-stm32.c:43:13: error: function declaration isn't a prototype
drivers/clocksource/timer-stm32.c: In function 'stm32_clock_event_set_mode':
drivers/clocksource/timer-stm32.c:47:3: error: type defaults to 'int' in declaration of '__mptr'
drivers/clocksource/timer-stm32.c:47:3: warning: initialization from incompatible pointer type
drivers/clocksource/timer-stm32.c:51:7: error: 'CLOCK_EVT_MODE_PERIODIC' undeclared (first use in this function)
drivers/clocksource/timer-stm32.c:51:7: note: each undeclared identifier is reported only once for each function it appears in
drivers/clocksource/timer-stm32.c:56:7: error: 'CLOCK_EVT_MODE_ONESHOT' undeclared (first use in this function)
Tighten up the dependencies to limit where it gets built by copying
the style of the Kconfig line for CLKSRC_EFM32 a few lines above.
Thomas Gleixner [Sun, 21 Jun 2015 14:21:50 +0000 (16:21 +0200)]
x86/hpet: Use proper hpet device number for MSI allocation
hpet_assign_irq() is called with hpet_device->num as "hardware
interrupt number", but hpet_device->num is initialized after the
interrupt has been assigned, so it's always 0. As a consequence only
the first MSI allocation succeeds, the following ones fail because the
"hardware interrupt number" already exists.
Move the initialization of dev->num and other fields before the call
to hpet_assign_irq(), which is the ordering before the offending
commit which introduced that regression.
The picoXcell hardware crypto accelerator driver was using an
older version of the clk framework, and not (un)preparing the
clock before enabling/disabling it. This change uses the handy
clk_prepare_enable function to interact with the current clk
framework correctly.
Herbert Xu [Fri, 19 Jun 2015 04:07:54 +0000 (12:07 +0800)]
crypto: nx - Check for bogus firmware properties
The nx driver reads two crucial paramters from the firmware for
each crypto algorithm, the maximum SG list length and byte limit.
Unfortunately those two parameters may be bogus, or worse they
may be absent altogether. When this happens the algorithms will
still register successfully but will fail when used or tested.
This patch adds checks to report any firmware entries which are
found to be bogus, and avoid registering algorithms which have
bogus parameters. A warning is also printed when an algorithm
is not registered because of this as there may have been no firmware
entries for it at all.
Linus Torvalds [Sat, 20 Jun 2015 20:54:22 +0000 (13:54 -0700)]
Merge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux
Pull drm fixes from Dave Airlie:
"A smattering of fixes,
mgag200:
don't accept modes that aren't aligned properly as hw can't do it
i915:
two regression fixes
radeon:
one query to allow userspace fixes
one oops fixer for older hw with new options enabled"
* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux:
drm/radeon: don't probe MST on hw we don't support it on
drm/radeon: Add RADEON_INFO_VA_UNMAP_WORKING query
drm/mgag200: Reject non-character-cell-aligned mode widths
Revert "drm/i915: Don't skip request retirement if the active list is empty"
drm/i915: Always reset vma->ggtt_view.pages cache on unbinding
Thomas Gleixner [Sat, 20 Jun 2015 10:05:40 +0000 (12:05 +0200)]
genirq: Remove bogus restriction in irq_move_mask_irq()
If an interrupt is marked with the no balancing flag, we still allow
setting the affinity for such an interrupt from the kernel itself, but
for interrupts which move the affinity from interrupt context via
irq_move_mask_irq() this runs into a check for the no balancing flag,
which in turn ends up with an endless storm of stack dumps because the
move pending flag is not reset.
Allow the move for interrupts which have the no balancing flag set and
clear the move pending bit before checking for interrupts with the per
cpu flag set.
Cisco has developed a new PCI HBA interface called sNIC, which stands for
SCSI NIC. This is a new storage feature supported on specialized network
adapter. The new PCI function provides a uniform host interface and abstracts
backend storage.
Kan Liang [Wed, 17 Jun 2015 13:51:11 +0000 (09:51 -0400)]
perf tools: Configurable per thread proc map processing time out
The time out to limit the individual proc map processing was hard code
to 500ms. This patch introduce a new option --proc-map-timeout to make
the time limit configurable.
Kan Liang [Wed, 17 Jun 2015 13:51:10 +0000 (09:51 -0400)]
perf tools: Add time out to force stop proc map processing
System wide sampling like 'perf top' or 'perf record -a' read all
threads /proc/xxx/maps before sampling. If there are any threads which
generating a keeping growing huge maps, perf will do infinite loop
during synthesizing. Nothing will be sampled.
This patch fixes this issue by adding per-thread timeout to force stop
this kind of endless proc map processing.
PERF_RECORD_MISC_PROC_MAP_PARSE_TIME_OUT is introduced to indicate that
the mmap record are truncated by time out. User will get warning
notification when truncated mmap records are detected.
Yannick Brosseau [Wed, 17 Jun 2015 23:41:10 +0000 (16:41 -0700)]
perf report: Fix sort__sym_cmp to also compare end of symbol
When using a map file from a JIT, due to memory reuse, we can obtain
multiple symbols with the same start address but a different length.
The symbols__find does check for the end so not doing it in
sort__sym_cmp was causing the hist_entry in the annotate part of a
report to match to the wrong entry, causing a fatal error.
perf hists browser: Do not exit when 'f' is pressed in 'report' mode
The 'f' hotkey is only used when in 'top', dynamic mode, to
enable/disable events, currently not making sense in the 'report',
static mode, where we can't go from showing the histogram entries
created from a perf.data file to adding more events after recreating the
evlist created from the perf.data file, albeit possible, this is not
implemented right now.
perf top: Replace CTRL+z with 'f' as hotkey for enable/disable events
I.e. 'freeze'/'unfreeze', this is because CTRL+z has a well known
action, i.e. suspend the app, perf needs to follow that convention, that
will be done on a separate patch, tho.
Martin Liška [Fri, 19 Jun 2015 19:10:43 +0000 (16:10 -0300)]
perf annotate: Display total number of samples with --show-total-period
To compare two records on an instruction base, with --show-total-period
option provided, display total number of samples that belong to a line
in assembly language.
New hot key 't' is introduced for 'perf annotate' TUI.
Adrian Hunter [Fri, 29 May 2015 13:33:30 +0000 (16:33 +0300)]
perf tools: Ensure thread-stack is flushed
The thread-stack represents a thread's current stack. When a thread
exits there can still be many functions on the stack e.g. exit() can be
called many levels deep, so all the callers will never return. To get
that information output, the thread-stack must be flushed.
Previously it was assumed the thread-stack would be flushed when the
struct thread was deleted. With thread ref-counting it is no longer
clear when that will be, if ever. So instead explicitly flush all the
thread-stacks at the end of a session.
Linus Torvalds [Fri, 19 Jun 2015 17:34:14 +0000 (07:34 -1000)]
Merge tag 'sound-4.1' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound
Pull sound fixes from Takashi Iwai:
"Nothing looks scary, just a few usual HD-audio regression fixes and
fixup, in addition to a minor Kconfig dependency fix for the old MIPS
drivers"
* tag 'sound-4.1' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound:
ALSA: hda - Fix unused label skip_i915
ALSA: hda - Fix noisy outputs on Dell XPS13 (2015 model)
ALSA: mips: let SND_SGI_O2 select SND_PCM
ALSA: hda - Fix audio crackles on Dell Latitude E7x40
ALSA: hda - adding a DAC/pin preference map for a HP Envy TS machine
Vladimir Murzin [Fri, 19 Jun 2015 14:28:03 +0000 (15:28 +0100)]
arm64: use private ratelimit state along with show_unhandled_signals
printk_ratelimit() shares the ratelimiting state with other callers what
may lead to scenarios where at the time we want to print out debug
information we already limited, so nothing appears in the dmesg - this
makes exception-trace quite poor helper in debugging.
Additionally, we have imbalance with some messages limited with global
ratelimit state and other messages limited with their private state
defined via pr_*_ratelimited().
To address this inconsistency show_unhandled_signals_ratelimited()
macro is introduced and caller sites are converted to use it.
projects / linux.git / log