summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* usb: gadget: pxa25x_udc cleanupArnd Bergmann2016-03-042-203/+275
| | | | | | | | | | | | | | | | | | | | | This removes the dependency on the mach/hardware.h header file from the pxa25x_udc driver after the register definitions were already unified in the previous patch. Following the model of pxa27x_udc (and basically all other drivers in the kernel), we define the register numbers as offsets from the register base address and use accessor functions to read/write them. For the moment, this still leaves the direct pointer dereference in place, instead of using readl/writel, so this patch should not be changing the behavior of the driver, other than using ioremap() on the platform resource to replace the hardcoded virtual address pointers. Acked-by: Robert Jarzmik <robert.jarzmik@free.fr> Signed-off-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Felipe Balbi <balbi@kernel.org>
* usb: gadget: pxa25x_udc: move register definitions from archArnd Bergmann2016-03-043-367/+155
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | ixp4xx and pxa25x both use this driver and provide a slightly different set of register definitions for it. Aside from that, the definition in the ixp4xx-regs.h header conflicts with the on in the pxa27x device driver when compile-testing that: In file included from ../drivers/usb/gadget/udc/pxa27x_udc.c:37:0: ../drivers/usb/gadget/udc/pxa27x_udc.h:26:0: warning: "UDCCR" redefined #define UDCCR 0x0000 /* UDC Control Register */ ^ In file included from ../arch/arm/mach-ixp4xx/include/mach/hardware.h:27:0, from ../arch/arm/mach-ixp4xx/include/mach/io.h:18, from ../arch/arm/include/asm/io.h:194, from ../include/linux/io.h:25, from ../include/linux/irq.h:24, from ../drivers/usb/gadget/udc/pxa27x_udc.c:23: ../arch/arm/mach-ixp4xx/include/mach/ixp4xx-regs.h:415:0: note: this is the location of the previous definition #define UDCCR IXP4XX_USB_REG(IXP4XX_USB_BASE_VIRT+0x0000) This addresses both issues by moving all the definitions into the pxa25x_udc driver itself. It turns out the only difference between them was 'UDCCS_IO_ROF', and that could well be a mistake when it was incorrectly copied from pxa25x to ixp4xx. Acked-by: Krzysztof Halasa <khalasa@piap.pl> Acked-by: Robert Jarzmik <robert.jarzmik@free.fr> Signed-off-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Felipe Balbi <balbi@kernel.org>
* usb: renesas_usbhs: Don't check CSSTS bit if peripheral modeYoshihiro Shimoda2016-03-041-4/+2
| | | | | | | | | | Since Some SoCs (e.g. R-Car Gen2) don't have the CSSTS bit in the pipectrl registers ({DCP,PIPEn}CTR) because such SoCs have peripheral mode only. So, this driver should not check the CSSTS bit if peripheral mode is running. Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com> Signed-off-by: Felipe Balbi <balbi@kernel.org>
* usb: renesas_usbhs: add R-Car Gen3 power controlYoshihiro Shimoda2016-03-045-5/+69
| | | | | | | | | | Since the usb2 phy driver for gen3 (phy-rcar-gen3-usb2) cannot access LPSTS and UGCTRL2 registers in the HSUSB module, this driver have to initialize the registers. So, this patch adds such handling code into rcar3.c. Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com> Signed-off-by: Felipe Balbi <balbi@kernel.org>
* usb: dwc3: Enable SuperSpeedPlusJohn Youn2016-03-041-0/+8
| | | | | | | | Enable SuperSpeedPlus by programming the DCFG.speed and after enumerating, set gadget->speed appropriately. Signed-off-by: John Youn <johnyoun@synopsys.com> Signed-off-by: Felipe Balbi <balbi@kernel.org>
* usb: dwc3: Update maximum_speed for SuperSpeedPlusJohn Youn2016-03-041-4/+13
| | | | | | | | | | | | | If the maximum_speed is not set, set it to a known value, either SuperSpeed or SuperSpeedPlus based on the type of controller we are using. If we are on DWC_usb31 controller, check the PHY interface to see if it is capable of SuperSpeedPlus. Also this check is moved after dwc3_core_init() so that we can check dwc->revision. Signed-off-by: John Youn <johnyoun@synopsys.com> Signed-off-by: Felipe Balbi <balbi@kernel.org>
* usb: dwc3: Update speed checks for SuperSpeedPlusJohn Youn2016-03-042-8/+14
| | | | | | | | Update various places where the speed is checked so that it takes into account SuperSpeedPlus properly. Signed-off-by: John Youn <johnyoun@synopsys.com> Signed-off-by: Felipe Balbi <balbi@kernel.org>
* usb: dwc3: Update register fields for SuperSpeedPlusJohn Youn2016-03-041-1/+4
| | | | | | | | Update various registers fields definitions for the DWC_usb31 controller for SuperSpeedPlus support. Signed-off-by: John Youn <johnyoun@synopsys.com> Signed-off-by: Felipe Balbi <balbi@kernel.org>
* usb: dwc3: DWC_usb31 controller checkJohn Youn2016-03-041-0/+6
| | | | | | | Add a convenience function to check if the controller is DWC_usb31. Signed-off-by: John Youn <johnyoun@synopsys.com> Signed-off-by: Felipe Balbi <balbi@kernel.org>
* usb: gadget: f_mass_storage: Enable SuperSpeedPlusJohn Youn2016-03-041-1/+1
| | | | | | | | | | | Enable superspeed plus configuration for the mass storage gadget. The mass storage function doesn't do anything special for SuperSpeedPlus. Just pass in the same SuperSpeed descriptors for SuperSpeedPlus. Signed-off-by: John Youn <johnyoun@synopsys.com> Signed-off-by: Felipe Balbi <balbi@kernel.org>
* usb: gadget: composite: Update debug message for SuperSpeedPlusJohn Youn2016-03-041-1/+2
| | | | | | | | Update the debug message reporting the speeds that a configuration supports for SuperSpeedPlus. Signed-off-by: John Youn <johnyoun@synopsys.com> Signed-off-by: Felipe Balbi <balbi@kernel.org>
* usb: gadget: composite: Configure the usb_ep for SuperSpeedPlusJohn Youn2016-03-041-1/+8
| | | | | | | | Configure the usb_ep using the SuperSpeedPlus descriptors if connected in SuperSpeedPlus. Signed-off-by: John Youn <johnyoun@synopsys.com> Signed-off-by: Felipe Balbi <balbi@kernel.org>
* usb: gadget: composite: Write SuperSpeedPlus config descriptorsJohn Youn2016-03-041-1/+5
| | | | | | | | Enable writing of SuperSpeedPlus descriptors for any SuperSpeedPlus capable configuration when connected in SuperSpeedPlus. Signed-off-by: John Youn <johnyoun@synopsys.com> Signed-off-by: Felipe Balbi <balbi@kernel.org>
* usb: gadget: composite: Add function to get descriptorsJohn Youn2016-03-041-21/+32
| | | | | | | | | There are a couple places in the code that get the function descriptors based on the speed. Move this lookup into a function call and add support to handle the SuperSpeedPlus descriptors as well. Signed-off-by: John Youn <johnyoun@synopsys.com> Signed-off-by: Felipe Balbi <balbi@kernel.org>
* usb: gadget: composite: Count configs for SuperSpeedPlusJohn Youn2016-03-041-1/+7
| | | | | | | | If enumerated in SuperSpeedPlus, count the configurations that support it. Signed-off-by: John Youn <johnyoun@synopsys.com> Signed-off-by: Felipe Balbi <balbi@kernel.org>
* usb: gadget: Update config for SuperSpeedPlusJohn Youn2016-03-043-0/+4
| | | | | | | | | | | | When a function is added to a configuration with usb_add_function(), the configuration speed flags are updated. These flags indicate for which speeds the configuration is valid for. This patch adds a flag in the configuration for SuperSpeedPlus and also updates this based on the existence of ssp_descriptors. Signed-off-by: John Youn <johnyoun@synopsys.com> Signed-off-by: Felipe Balbi <balbi@kernel.org>
* usb: gadget: Update function for SuperSpeedPlusJohn Youn2016-03-042-0/+11
| | | | | | | | | | Add a ssp_descriptors member to struct usb_function and handle the initialization and cleanup of it. This holds the SuperSpeedPlus descriptors for a function that supports SuperSpeedPlus. This is added by usb_assign_descriptors(). Signed-off-by: John Youn <johnyoun@synopsys.com> Signed-off-by: Felipe Balbi <balbi@kernel.org>
* usb: gadget: Update usb_assign_descriptors for SuperSpeedPlusJohn Youn2016-03-0419-19/+24
| | | | | | | | | | | | Add the 'ssp_descriptors' parameter to the usb_assign_descriptors() function. This allows a function driver to add descriptors for SuperSpeedPlus speeds if it supports it. Also update all uses of this function in the gadget subsystem so that they pass NULL for the ssp_descriptors parameters. Signed-off-by: John Youn <johnyoun@synopsys.com> Signed-off-by: Felipe Balbi <balbi@kernel.org>
* usb: gadget: composite: Return SSP Dev Cap descriptorJohn Youn2016-03-041-0/+42
| | | | | | | | | | | If a gadget supports SuperSpeedPlus or higher speeds, return a SuperSpeedPlus USB Device Capability descriptor. Currently this implementation returns a fixed descriptor with typical values set. Signed-off-by: John Youn <johnyoun@synopsys.com> Signed-off-by: Felipe Balbi <balbi@kernel.org>
* usb: gadget: composite: Return bcdUSB 0x0310John Youn2016-03-041-1/+1
| | | | | | | | | | | The USB 3.1 specification replaces the USB 3.0 specification and all new devices that are running at SuperSpeed or higher speeds must report a bcdUSB of 0x0310. Refer to USB 3.1 Specification, Revision 1.0, Section 9.6. Signed-off-by: John Youn <johnyoun@synopsys.com> Signed-off-by: Felipe Balbi <balbi@kernel.org>
* usb: gadget: Add gadget_is_superspeed_plus()John Youn2016-03-041-0/+10
| | | | | | | Add a function to check for SuperSpeedPlus capable gadgets. Signed-off-by: John Youn <johnyoun@synopsys.com> Signed-off-by: Felipe Balbi <balbi@kernel.org>
* usb: ch9: Add size macro for SSP dev cap descriptorJohn Youn2016-03-041-0/+6
| | | | | | | | | | | | | | The SuperspeedPlus Device Capability Descriptor has a variable size depending on the number of sublink speed attributes. This patch adds a macro to calculate that size. The macro takes one argument, the Sublink Speed Attribute Count (SSAC) as reported by the descriptor in bmAttributes[4:0]. See USB 3.1 9.6.2.5, Table 9-19. Signed-off-by: John Youn <johnyoun@synopsys.com> Signed-off-by: Felipe Balbi <balbi@kernel.org>
* Merge tag 'usb-ci-v4.6-rc1' of ↵Greg Kroah-Hartman2016-03-024-5/+35
|\ | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/peter.chen/usb into usb-next Peter writes: - Add platform interface to choose ttctrl.ttha - Some tiny improvements
| * usb: chipidea: imx: avoid EPROBE_DEFER printed as errorStefan Agner2016-02-291-3/+3
| | | | | | | | | | | | | | | | | | Avoid printing an error if adding the device failes with return value EPROBE_DEFFER. This may happen e.g. due to missing GPIO for the vbus-supply regulator. Signed-off-by: Stefan Agner <stefan@agner.ch> Signed-off-by: Peter Chen <peter.chen@nxp.com>
| * Revert "usb: chipidea: imx: enable CI_HDRC_SET_NON_ZERO_TTHA"Peter Chen2016-02-291-1/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | This reverts commit e765bfb73ff7. In the most of cases, we only use one transaction per frame and the frame rate may be high, If the platforms want to support multiple transactions but less frame rate cases like [1] and [2], it can set "non-zero-ttctrl-ttha" at dts. [1] http://www.spinics.net/lists/linux-usb/msg123125.html [2] http://www.spinics.net/lists/linux-usb/msg118679.html Signed-off-by: Peter Chen <peter.chen@nxp.com>
| * doc: usb: ci-hdrc-usb2: add property non-zero-ttctrl-tthaPeter Chen2016-02-291-0/+16
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | If this property is not set, the max packet size is 1023 bytes, and if the total of packet size for pervious transactions are more than 256 bytes, it can't accept any transactions within this frame. The use case is single transaction, but higher frame rate. If this property is set, the max packet size is 188 bytes, it can handle more transactions than above case, it can accept transactions until it considers the left room size within frame is less than 188 bytes, software needs to make sure it does not send more than 90% maximum_periodic_data_per_frame. The use case is multiple transactions, but less frame rate. Signed-off-by: Peter Chen <peter.chen@nxp.com>
| * usb: chipidea: add system interface for ttctrl.tthaPeter Chen2016-02-291-0/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In chipidea IP RTL, there is a very limited design for siTD, the detail like below: There is no Max Packet Size at siTD, so it uses one constant for both Max Packet Size for packet and the packet size for the last transaction when considering schedule. If the ttctrl.ttha does not match against Hub Address field in siTD, this constant is 188 bytes, else this constant is 1023 bytes. If the ttctrl.ttha is non-zero value, RTL will use 188 as this constant, so it will lose the data if the packet size is larger than 188 bytes, eg, if we playback a wav which format is 48khz, 16 bits, 2 channels, the packet size will be 192bytes, but the controller will only send 188 bytes for this packet, the noise will be heared using USB audio card. The use case is single transaction, but higher frame rate. If the ttctr.ttha is zero value, we can send 1023 bytes within one transaction, but the controller will not accept the coming tranaction if it considers the schedule time is less than 1023 bytes. So the limitation is we can't schedule as many as transactions within frame. If the total bytes is already 256 bytes for previous transactions within frame, it can't accept another transaction. The use case is multiple transactions, but less frame rate. Signed-off-by: Peter Chen <peter.chen@nxp.com>
| * usb: chipidea: udc: remove unused value assignmentLi Jun2016-02-291-1/+0
| | | | | | | | | | | | | | | | retval is assigned to be -EOVERFLOW but is overwritten later before it's used, remove this unused value assignment. Signed-off-by: Li Jun <jun.li@nxp.com> Signed-off-by: Peter Chen <peter.chen@nxp.com>
| * dt-bindings: ci-hdrc-usb2: add missing compatiblesStefan Wahren2016-02-291-0/+7
| | | | | | | | | | | | | | This patch adds the missing compatible strings from ci_hdrc_imx. Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com> Signed-off-by: Peter Chen <peter.chen@nxp.com>
| * usb: chipidea: add CI_HDRC_TURN_VBUS_EARLY_ON for imx23Stefan Wahren2016-02-291-0/+6
| | | | | | | | | | | | | | | | | | | | Until now the imx23 uses the imx27 platform flag. But the imx23 needs the flag CI_HDRC_TURN_VBUS_EARLY_ON, too. So fix this by adding a separate platform flag. Suggested-by: Peter Chen <peter.chen@nxp.com> Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com> Signed-off-by: Peter Chen <peter.chen@nxp.com>
* | Merge 4.5-rc6 into usb-nextGreg Kroah-Hartman2016-03-02537-4374/+7314
|\ \ | |/ |/| | | | | | | We want the USB fixes in here as well. Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
| * Linux 4.5-rc6v4.5-rc6Linus Torvalds2016-02-281-1/+1
| |
| * Merge branch 'perf-urgent-for-linus' of ↵Linus Torvalds2016-02-282-131/+244
| |\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip Pull perf fixes from Thomas Gleixner: "A rather largish series of 12 patches addressing a maze of race conditions in the perf core code from Peter Zijlstra" * 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: perf: Robustify task_function_call() perf: Fix scaling vs. perf_install_in_context() perf: Fix scaling vs. perf_event_enable() perf: Fix scaling vs. perf_event_enable_on_exec() perf: Fix ctx time tracking by introducing EVENT_TIME perf: Cure event->pending_disable race perf: Fix race between event install and jump_labels perf: Fix cloning perf: Only update context time when active perf: Allow perf_release() with !event->ctx perf: Do not double free perf: Close install vs. exit race
| | * perf: Robustify task_function_call()Peter Zijlstra2016-02-251-20/+20
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Since there is no serialization between task_function_call() doing task_curr() and the other CPU doing context switches, we could end up not sending an IPI even if we had to. And I'm not sure I still buy my own argument we're OK. Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: dvyukov@google.com Cc: eranian@google.com Cc: oleg@redhat.com Cc: panand@redhat.com Cc: sasha.levin@oracle.com Cc: vince@deater.net Link: http://lkml.kernel.org/r/20160224174948.340031200@infradead.org Signed-off-by: Ingo Molnar <mingo@kernel.org>
| | * perf: Fix scaling vs. perf_install_in_context()Peter Zijlstra2016-02-251-45/+70
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Completely reworks perf_install_in_context() (again!) in order to ensure that there will be no ctx time hole between add_event_to_ctx() and any potential ctx_sched_in(). Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: dvyukov@google.com Cc: eranian@google.com Cc: oleg@redhat.com Cc: panand@redhat.com Cc: sasha.levin@oracle.com Cc: vince@deater.net Link: http://lkml.kernel.org/r/20160224174948.279399438@infradead.org Signed-off-by: Ingo Molnar <mingo@kernel.org>
| | * perf: Fix scaling vs. perf_event_enable()Peter Zijlstra2016-02-251-19/+23
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Similar to the perf_enable_on_exec(), ensure that event timings are consistent across perf_event_enable(). Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: dvyukov@google.com Cc: eranian@google.com Cc: oleg@redhat.com Cc: panand@redhat.com Cc: sasha.levin@oracle.com Cc: vince@deater.net Link: http://lkml.kernel.org/r/20160224174948.218288698@infradead.org Signed-off-by: Ingo Molnar <mingo@kernel.org>
| | * perf: Fix scaling vs. perf_event_enable_on_exec()Peter Zijlstra2016-02-251-0/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The recent commit 3e349507d12d ("perf: Fix perf_enable_on_exec() event scheduling") caused this by moving task_ctx_sched_out() from before __perf_event_mask_enable() to after it. The overlooked consequence of that change is that task_ctx_sched_out() would update the ctx time fields, and now __perf_event_mask_enable() uses stale time. In order to fix this, explicitly stop our context's time before enabling the event(s). Reported-by: Oleg Nesterov <oleg@redhat.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: dvyukov@google.com Cc: eranian@google.com Cc: panand@redhat.com Cc: sasha.levin@oracle.com Cc: vince@deater.net Fixes: 3e349507d12d ("perf: Fix perf_enable_on_exec() event scheduling") Link: http://lkml.kernel.org/r/20160224174948.159242158@infradead.org Signed-off-by: Ingo Molnar <mingo@kernel.org>
| | * perf: Fix ctx time tracking by introducing EVENT_TIMEPeter Zijlstra2016-02-251-12/+30
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Currently any ctx_sched_in() call will re-start the ctx time tracking, this means that calls like: ctx_sched_in(.event_type = EVENT_PINNED); ctx_sched_in(.event_type = EVENT_FLEXIBLE); will have a hole in their ctx time tracking. This is likely harmless but can confuse things a little. By adding EVENT_TIME, we can have the first ctx_sched_in() (is_active: 0 -> !0) start the time and any further ctx_sched_in() will leave the timestamps alone. Secondly, this allows for an early disable like: ctx_sched_out(.event_type = EVENT_TIME); which would update the ctx time (if the ctx is active) and any further calls to ctx_sched_out() would not further modify the ctx time. For ctx_sched_in() any 0 -> !0 transition will automatically include EVENT_TIME. For ctx_sched_out(), any transition that clears EVENT_ALL will automatically clear EVENT_TIME. These two rules ensure that under normal circumstances we need not bother with EVENT_TIME and get natural ctx time behaviour. Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: dvyukov@google.com Cc: eranian@google.com Cc: oleg@redhat.com Cc: panand@redhat.com Cc: sasha.levin@oracle.com Cc: vince@deater.net Link: http://lkml.kernel.org/r/20160224174948.100446561@infradead.org Signed-off-by: Ingo Molnar <mingo@kernel.org>
| | * perf: Cure event->pending_disable racePeter Zijlstra2016-02-251-3/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Because event_sched_out() checks event->pending_disable _before_ actually disabling the event, it can happen that the event fires after it checks but before it gets disabled. This would leave event->pending_disable set and the queued irq_work will try and process it. However, if the event trigger was during schedule(), the event might have been de-scheduled by the time the irq_work runs, and perf_event_disable_local() will fail. Fix this by checking event->pending_disable _after_ we call event->pmu->del(). This depends on the latter being a compiler barrier, such that the compiler does not lift the load and re-creates the problem. Tested-by: Alexander Shishkin <alexander.shishkin@linux.intel.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: dvyukov@google.com Cc: eranian@google.com Cc: oleg@redhat.com Cc: panand@redhat.com Cc: sasha.levin@oracle.com Cc: vince@deater.net Link: http://lkml.kernel.org/r/20160224174948.040469884@infradead.org Signed-off-by: Ingo Molnar <mingo@kernel.org>
| | * perf: Fix race between event install and jump_labelsPeter Zijlstra2016-02-252-11/+44
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | perf_install_in_context() relies upon the context switch hooks to have scheduled in events when the IPI misses its target -- after all, if the task has moved from the CPU (or wasn't running at all), it will have to context switch to run elsewhere. This however doesn't appear to be happening. It is possible for the IPI to not happen (task wasn't running) only to later observe the task running with an inactive context. The only possible explanation is that the context switch hooks are not called. Therefore put in a sync_sched() after toggling the jump_label to guarantee all CPUs will have them enabled before we install an event. A simple if (0->1) sync_sched() will not in fact work, because any further increment can race and complete before the sync_sched(). Therefore we must jump through some hoops. Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: dvyukov@google.com Cc: eranian@google.com Cc: oleg@redhat.com Cc: panand@redhat.com Cc: sasha.levin@oracle.com Cc: vince@deater.net Link: http://lkml.kernel.org/r/20160224174947.980211985@infradead.org Signed-off-by: Ingo Molnar <mingo@kernel.org>
| | * perf: Fix cloningPeter Zijlstra2016-02-252-15/+15
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Alexander reported that when the 'original' context gets destroyed, no new clones happen. This can happen irrespective of the ctx switch optimization, any task can die, even the parent, and we want to continue monitoring the task hierarchy until we either close the event or no tasks are left in the hierarchy. perf_event_init_context() will attempt to pin the 'parent' context during clone(). At that point current is the parent, and since current cannot have exited while executing clone(), its context cannot have passed through perf_event_exit_task_context(). Therefore perf_pin_task_context() cannot observe ctx->task == TASK_TOMBSTONE. However, since inherit_event() does: if (parent_event->parent) parent_event = parent_event->parent; it looks at the 'original' event when it does: is_orphaned_event(). This can return true if the context that contains the this event has passed through perf_event_exit_task_context(). And thus we'll fail to clone the perf context. Fix this by adding a new state: STATE_DEAD, which is set by perf_release() to indicate that the filedesc (or kernel reference) is dead and there are no observers for our data left. Only for STATE_DEAD will is_orphaned_event() be true and inhibit cloning. STATE_EXIT is otherwise preserved such that is_event_hup() remains functional and will report when the observed task hierarchy becomes empty. Reported-by: Alexander Shishkin <alexander.shishkin@linux.intel.com> Tested-by: Alexander Shishkin <alexander.shishkin@linux.intel.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: dvyukov@google.com Cc: eranian@google.com Cc: oleg@redhat.com Cc: panand@redhat.com Cc: sasha.levin@oracle.com Cc: vince@deater.net Fixes: c6e5b73242d2 ("perf: Synchronously clean up child events") Link: http://lkml.kernel.org/r/20160224174947.919845295@infradead.org Signed-off-by: Ingo Molnar <mingo@kernel.org>
| | * perf: Only update context time when activePeter Zijlstra2016-02-251-6/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: dvyukov@google.com Cc: eranian@google.com Cc: oleg@redhat.com Cc: panand@redhat.com Cc: sasha.levin@oracle.com Cc: vince@deater.net Link: http://lkml.kernel.org/r/20160224174947.860690919@infradead.org Signed-off-by: Ingo Molnar <mingo@kernel.org>
| | * perf: Allow perf_release() with !event->ctxPeter Zijlstra2016-02-251-3/+13
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In the err_file: fput(event_file) case, the event will not yet have been attached to a context. However perf_release() does assume it has been. Cure this. Tested-by: Alexander Shishkin <alexander.shishkin@linux.intel.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: dvyukov@google.com Cc: eranian@google.com Cc: oleg@redhat.com Cc: panand@redhat.com Cc: sasha.levin@oracle.com Cc: vince@deater.net Link: http://lkml.kernel.org/r/20160224174947.793996260@infradead.org Signed-off-by: Ingo Molnar <mingo@kernel.org>
| | * perf: Do not double freePeter Zijlstra2016-02-251-1/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In case of: err_file: fput(event_file), we'll end up calling perf_release() which in turn will free the event. Do not then free the event _again_. Tested-by: Alexander Shishkin <alexander.shishkin@linux.intel.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: dvyukov@google.com Cc: eranian@google.com Cc: oleg@redhat.com Cc: panand@redhat.com Cc: sasha.levin@oracle.com Cc: vince@deater.net Link: http://lkml.kernel.org/r/20160224174947.697350349@infradead.org Signed-off-by: Ingo Molnar <mingo@kernel.org>
| | * perf: Close install vs. exit racePeter Zijlstra2016-02-251-9/+26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Consider the following scenario: CPU0 CPU1 ctx = find_get_ctx(); perf_event_exit_task_context() mutex_lock(&ctx->mutex); perf_install_in_context(ctx, ...); /* NO-OP */ mutex_unlock(&ctx->mutex); ... perf_release() WARN_ON_ONCE(event->state != STATE_EXIT); Since the event doesn't pass through perf_remove_from_context() because perf_install_in_context() NO-OPs because the ctx is dead, and perf_event_exit_task_context() will not observe the event because its not attached yet, the event->state will not be set. Solve this by revalidating ctx->task after we acquire ctx->mutex and failing the event creation as a whole. Tested-by: Alexander Shishkin <alexander.shishkin@linux.intel.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: dvyukov@google.com Cc: eranian@google.com Cc: oleg@redhat.com Cc: panand@redhat.com Cc: sasha.levin@oracle.com Cc: vince@deater.net Link: http://lkml.kernel.org/r/20160224174947.626853419@infradead.org Signed-off-by: Ingo Molnar <mingo@kernel.org>
| * | Merge branch 'x86-urgent-for-linus' of ↵Linus Torvalds2016-02-285-7/+15
| |\ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip Pull x86 fixes from Thomas Gleixner: "This update contains: - Hopefully the last ASM CLAC fixups - A fix for the Quark family related to the IMR lock which makes kexec work again - A off-by-one fix in the MPX code. Ironic, isn't it? - A fix for X86_PAE which addresses once more an unsigned long vs phys_addr_t hickup" * 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: x86/mpx: Fix off-by-one comparison with nr_registers x86/mm: Fix slow_virt_to_phys() for X86_PAE again x86/entry/compat: Add missing CLAC to entry_INT80_32 x86/entry/32: Add an ASM_CLAC to entry_SYSENTER_32 x86/platform/intel/quark: Change the kernel's IMR lock bit to false
| | * | x86/mpx: Fix off-by-one comparison with nr_registersColin Ian King2016-02-261-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In the unlikely event that regno == nr_registers then we get an array overrun on regoff because the invalid register check is currently off-by-one. Fix this with a check that regno is >= nr_registers instead. Detected with static analysis using CoverityScan. Fixes: fcc7ffd67991 "x86, mpx: Decode MPX instruction to get bound violation information" Signed-off-by: Colin Ian King <colin.king@canonical.com> Acked-by: Dave Hansen <dave.hansen@linux.intel.com> Cc: Borislav Petkov <bp@alien8.de> Cc: "Kirill A . Shutemov" <kirill.shutemov@linux.intel.com> Cc: stable@vger.kernel.org Link: http://lkml.kernel.org/r/1456512931-3388-1-git-send-email-colin.king@canonical.com Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
| | * | x86/mm: Fix slow_virt_to_phys() for X86_PAE againDexuan Cui2016-02-251-4/+10
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | "d1cd12108346: x86, pageattr: Prevent overflow in slow_virt_to_phys() for X86_PAE" was unintentionally removed by the recent "34437e67a672: x86/mm: Fix slow_virt_to_phys() to handle large PAT bit". And, the variable 'phys_addr' was defined as "unsigned long" by mistake -- it should be "phys_addr_t". As a result, Hyper-V network driver in 32-PAE Linux guest can't work again. Fixes: commit 34437e67a672: "x86/mm: Fix slow_virt_to_phys() to handle large PAT bit" Signed-off-by: Dexuan Cui <decui@microsoft.com> Reviewed-by: Toshi Kani <toshi.kani@hpe.com> Cc: olaf@aepfle.de Cc: gregkh@linuxfoundation.org Cc: jasowang@redhat.com Cc: driverdev-devel@linuxdriverproject.org Cc: linux-mm@kvack.org Cc: apw@canonical.com Cc: Andrew Morton <akpm@linux-foundation.org> Cc: K. Y. Srinivasan <kys@microsoft.com> Cc: Haiyang Zhang <haiyangz@microsoft.com> Link: http://lkml.kernel.org/r/1456394292-9030-1-git-send-email-decui@microsoft.com Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
| | * | x86/entry/compat: Add missing CLAC to entry_INT80_32Andy Lutomirski2016-02-251-0/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This doesn't seem to fix a regression -- I don't think the CLAC was ever there. I double-checked in a debugger: entries through the int80 gate do not automatically clear AC. Stable maintainers: I can provide a backport to 4.3 and earlier if needed. This needs to be backported all the way to 3.10. Reported-by: Brian Gerst <brgerst@gmail.com> Signed-off-by: Andy Lutomirski <luto@kernel.org> Cc: Andy Lutomirski <luto@amacapital.net> Cc: Borislav Petkov <bp@alien8.de> Cc: Denys Vlasenko <dvlasenk@redhat.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: <stable@vger.kernel.org> # v3.10 and later Fixes: 63bcff2a307b ("x86, smap: Add STAC and CLAC instructions to control user space access") Link: http://lkml.kernel.org/r/b02b7e71ae54074be01fc171cbd4b72517055c0e.1456345086.git.luto@kernel.org Signed-off-by: Ingo Molnar <mingo@kernel.org>
| | * | x86/entry/32: Add an ASM_CLAC to entry_SYSENTER_32Andy Lutomirski2016-02-241-0/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Both before and after 5f310f739b4c ("x86/entry/32: Re-implement SYSENTER using the new C path"), we relied on a uaccess very early in the SYSENTER path to clear AC. After that change, though, we can potentially make it all the way into C code with AC set, which enlarges the attack surface for SMAP bypass by doing SYSENTER with AC set. Strengthen the SMAP protection by addding the missing ASM_CLAC right at the beginning. Signed-off-by: Andy Lutomirski <luto@kernel.org> Cc: Andy Lutomirski <luto@amacapital.net> Cc: Borislav Petkov <bp@alien8.de> Cc: Brian Gerst <brgerst@gmail.com> Cc: Denys Vlasenko <dvlasenk@redhat.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Link: http://lkml.kernel.org/r/3e36be110724896e32a4a1fe73bacb349d3cba94.1456262295.git.luto@kernel.org Signed-off-by: Ingo Molnar <mingo@kernel.org>
generated by cgit v1.2.3 (git 2.39.1) at 2025-01-04 07:13:37 +0100