[PATCH v3 08/20] coresight: allow etm4x to be built as a module
Mathieu Poirier
mathieu.poirier at linaro.org
Tue Jul 21 10:55:07 EDT 2020
Hi Tingwei,
On Sat, Jul 18, 2020 at 06:38:08PM +0100, Mike Leach wrote:
> Hi,
>
> That fix sorts the problem on my board.
> However, be aware that the etmv4 PM code changes somewhat due to this
> patch [1] that is currently on the linux master branch and will appear
> in 5.8-rc6, and there are further new coresight patches in Mathieu's
> coresight/next and linux-next.
>
> It may be better to rebase onto coresight/next for the next set.
>
As Mike suggested please rebase on coresight/next and send another revision.
Thanks,
Mathieu
> Regards
>
> Mike
>
> [1] 9b6a3f3633a5 coresight: etmv4: Fix CPU power management setup in
> probe() function
>
>
> On Sat, 18 Jul 2020 at 04:25, <tingwei at codeaurora.org> wrote:
> >
> > Hi Mike,
> >
> > Thanks for reporting this. This is a good catch.
> >
> > I was testing on db845c but I didn't encounter this issue with
> > multiple module load/unload test. I guess it's due to timing difference.
> > The issue is etm4_cpu_pm_notify() is removed from cpu pm notification
> > after last etm4 device is removed. There's racing condition when some
> > etm4 device say etm0 is removed but etm4_cpu_pm_notify() is still
> > called. I reproduced this issue by manually unbind etm4x device.
> >
> > The fix is clean etmdrvdata[drvdata->cpu] in etm4_remove() so
> > etm4_cpu_pm_notify() is gracefully returned. I've verified this on
> > db845c with manually unbind etm4x device test. Do you mind test
> > on your db410 as well? I'll put it into v4.
> >
> > @@ -1542,6 +1543,7 @@ static int __exit etm4_remove(struct amba_device
> > *adev)
> >
> > etm_perf_symlink(drvdata->csdev, false);
> >
> > + etmdrvdata[drvdata->cpu] = NULL;
> > if (--etm4_count == 0) {
> > etm4_cpu_pm_unregister();
> >
> >
> >
> > On 2020-07-18 01:05, Mike Leach wrote:
> > > Hi Tingwei,
> > >
> > > Couldn't see 00/20 cover note for this set, so reporting this here as
> > > it relates to etmv4 in part.
> > >
> > > The following sequence:-
> > > (There is a load dependency here of course, if I try to load ETMv4
> > > without the coresight core, then I get a lot of missing symbol error
> > > messages - I assume this is expected behaviour?)
> >
> > Coresight_etm4x module depends on coresight module, so it's expected
> > behavior.
> >
> > >
> > > root at linaro-developer:/home/linaro/cs-mods# insmod coresight.ko
> > > root at linaro-developer:/home/linaro/cs-mods# insmod coresight-etm4x.ko
> > >
> > > correctly loads the coresight core then ETMv4 module.
> > >
> > > [ 1208.214674] coresight etm0: CPU0: ETM v4.0 initialized
> > > [ 1208.215534] coresight etm1: CPU1: ETM v4.0 initialized
> > > [ 1208.221020] coresight etm2: CPU2: ETM v4.0 initialized
> > > [ 1208.224757] coresight etm3: CPU3: ETM v4.0 initialized
> > >
> > > However, if I then unload the ETMv4 module:-
> > >
> > > root at linaro-developer:/home/linaro/cs-mods# rmmod coresight-etm4x.ko
> > >
> > > I get a crash:-
> > >
> > > [ 1215.963689] ------------[ cut here ]------------
> > > [ 1215.963741] WARNING: CPU: 3 PID: 0 at
> > > drivers/hwtracing/coresight/coresight-etm4x-core.c:1364
> > > etm4_cpu_pm_notify+0xad8/0xb48 [coresight_etm4x]
> > > [ 1215.967373] Modules linked in: coresight_etm4x(-) coresight [last
> > > unloaded: coresight]
> > > [ 1215.979960] CPU: 3 PID: 0 Comm: swapper/3 Tainted: G W
> > > 5.8.0-rc5cs-modscs-mods-00020-gc03fe910680d #282
> > > [ 1215.987856] Hardware name: Qualcomm Technologies, Inc. APQ 8016 SBC
> > > (DT)
> > > [ 1215.998531] pstate: 80000085 (Nzcv daIf -PAN -UAO BTYPE=--)
> > > [ 1216.005396] pc : etm4_cpu_pm_notify+0xad8/0xb48 [coresight_etm4x]
> > > [ 1216.010687] lr : notifier_call_chain+0x5c/0xa0
> > > [ 1216.016926] sp : ffff000009fe7d50
> > > [ 1216.021265] x29: ffff000009fe7d50 x28: 0000000000000000
> > > [ 1216.024653] x27: 0000011b1d1464b5 x26: 0000000000000001
> > > [ 1216.030035] x25: ffff00003c79f888 x24: 0000000000000001
> > > [ 1216.035330] x23: 0000000000000000 x22: 0000000000000000
> > > [ 1216.040626] x21: 0000000000000000 x20: 0000000000000000
> > > [ 1216.045921] x19: ffff000038396480 x18: 0000000000000000
> > > [ 1216.051216] x17: 0000000000000000 x16: 0000000000000000
> > > [ 1216.056510] x15: 0000000000000000 x14: ffffffffffffffff
> > > [ 1216.061806] x13: ffffffffffffffff x12: 0000000000000000
> > > [ 1216.067101] x11: 0000000000000000 x10: 00000000000009b0
> > > [ 1216.072396] x9 : ffff000009fe7e70 x8 : ffff000009fd4010
> > > [ 1216.077690] x7 : 0000000000000000 x6 : 0000000023c2c562
> > > [ 1216.082986] x5 : 00ffffffffffffff x4 : 0000000000000003
> > > [ 1216.088281] x3 : ffff000038396880 x2 : ffff800008d8d280
> > > [ 1216.093576] x1 : ffff80002b1aa000 x0 : 000000005f454c42
> > > [ 1216.098873] Call trace:
> > > [ 1216.104178] etm4_cpu_pm_notify+0xad8/0xb48 [coresight_etm4x]
> > > [ 1216.106343] notifier_call_chain+0x5c/0xa0
> > > [ 1216.112242] __atomic_notifier_call_chain+0x48/0x60
> > > [ 1216.116236] cpu_pm_notify+0x44/0x70
> > > [ 1216.121008] cpu_pm_enter+0x3c/0x80
> > > [ 1216.124829] psci_enter_domain_idle_state+0x38/0xa8
> > > [ 1216.128040] cpuidle_enter_state+0x88/0x478
> > > [ 1216.132900] cpuidle_enter+0x44/0x58
> > > [ 1216.137069] call_cpuidle+0x40/0x70
> > > [ 1216.140888] do_idle+0x1e0/0x248
> > > [ 1216.144099] cpu_startup_entry+0x28/0x98
> > > [ 1216.147574] secondary_start_kernel+0x1a0/0x1f8
> > > [ 1216.151476] ---[ end trace 97bcd7b8bdd7f9a7 ]---
> > >
> > > This happens each time, though the CPU number can change.
> > > Test platform is Dragonboard DB410, linux 5.8-rc5 + this patch set.
> > >
> > > Regards
> > >
> > > Mike
> > >
> > >
> > > On Fri, 17 Jul 2020 at 06:48, Tingwei Zhang <tingwei at codeaurora.org>
> > > wrote:
> > >>
> > >> From: Kim Phillips <kim.phillips at arm.com>
> > >>
> > >> Allow to build coresight-etm4x as a module, for ease of development.
> > >>
> > >> - Kconfig becomes a tristate, to allow =m
> > >> - append -core to source file name to allow module to
> > >> be called coresight-etm4x by the Makefile
> > >> - add an etm4_remove function, for module unload
> > >> - add a MODULE_DEVICE_TABLE for autoloading on boot
> > >>
> > >> Cc: Mathieu Poirier <mathieu.poirier at linaro.org>
> > >> Cc: Leo Yan <leo.yan at linaro.org>
> > >> Cc: Alexander Shishkin <alexander.shishkin at linux.intel.com>
> > >> Cc: Randy Dunlap <rdunlap at infradead.org>
> > >> Cc: Suzuki K Poulose <Suzuki.Poulose at arm.com>
> > >> Cc: Greg Kroah-Hartman <gregkh at linuxfoundation.org>
> > >> Cc: Russell King <linux at armlinux.org.uk>
> > >> Signed-off-by: Kim Phillips <kim.phillips at arm.com>
> > >> Signed-off-by: Tingwei Zhang <tingwei at codeaurora.org>
> > >> ---
> > >> drivers/hwtracing/coresight/Kconfig | 5 ++-
> > >> drivers/hwtracing/coresight/Makefile | 4 +--
> > >> ...resight-etm4x.c => coresight-etm4x-core.c} | 31
> > >> ++++++++++++++++++-
> > >> 3 files changed, 36 insertions(+), 4 deletions(-)
> > >> rename drivers/hwtracing/coresight/{coresight-etm4x.c =>
> > > coresight-etm4x-core.c} (98%)
> > >>
> > >> diff --git a/drivers/hwtracing/coresight/Kconfig
> > > b/drivers/hwtracing/coresight/Kconfig
> > >> index 8fd9fd139cf3..d6e107bbd30b 100644
> > >> --- a/drivers/hwtracing/coresight/Kconfig
> > >> +++ b/drivers/hwtracing/coresight/Kconfig
> > >> @@ -78,7 +78,7 @@ config CORESIGHT_SOURCE_ETM3X
> > >> module will be called coresight-etm3x.
> > >>
> > >> config CORESIGHT_SOURCE_ETM4X
> > >> - bool "CoreSight Embedded Trace Macrocell 4.x driver"
> > >> + tristate "CoreSight Embedded Trace Macrocell 4.x driver"
> > >> depends on ARM64
> > >> select CORESIGHT_LINKS_AND_SINKS
> > >> select PID_IN_CONTEXTIDR
> > >> @@ -88,6 +88,9 @@ config CORESIGHT_SOURCE_ETM4X
> > >> for instruction level tracing. Depending on the implemented
> > > version
> > >> data tracing may also be available.
> > >>
> > >> + To compile this driver as a module, choose M here: the
> > >> + module will be called coresight-etm4x.
> > >> +
> > >> config CORESIGHT_STM
> > >> tristate "CoreSight System Trace Macrocell driver"
> > >> depends on (ARM && !(CPU_32v3 || CPU_32v4 || CPU_32v4T)) ||
> > > ARM64
> > >> diff --git a/drivers/hwtracing/coresight/Makefile
> > > b/drivers/hwtracing/coresight/Makefile
> > >> index d619cfd0abd8..271dc255454f 100644
> > >> --- a/drivers/hwtracing/coresight/Makefile
> > >> +++ b/drivers/hwtracing/coresight/Makefile
> > >> @@ -14,8 +14,8 @@ obj-$(CONFIG_CORESIGHT_LINKS_AND_SINKS) +=
> > > coresight-funnel.o \
> > >> obj-$(CONFIG_CORESIGHT_SOURCE_ETM3X) += coresight-etm3x.o
> > >> coresight-etm3x-y := coresight-etm3x-core.o coresight-etm-cp14.o \
> > >> coresight-etm3x-sysfs.o
> > >> -obj-$(CONFIG_CORESIGHT_SOURCE_ETM4X) += coresight-etm4x.o \
> > >> - coresight-etm4x-sysfs.o
> > >> +obj-$(CONFIG_CORESIGHT_SOURCE_ETM4X) += coresight-etm4x.o
> > >> +coresight-etm4x-y := coresight-etm4x-core.o coresight-etm4x-sysfs.o
> > >> obj-$(CONFIG_CORESIGHT_STM) += coresight-stm.o
> > >> obj-$(CONFIG_CORESIGHT_CPU_DEBUG) += coresight-cpu-debug.o
> > >> obj-$(CONFIG_CORESIGHT_CATU) += coresight-catu.o
> > >> diff --git a/drivers/hwtracing/coresight/coresight-etm4x.c
> > > b/drivers/hwtracing/coresight/coresight-etm4x-core.c
> > >> similarity index 98%
> > >> rename from drivers/hwtracing/coresight/coresight-etm4x.c
> > >> rename to drivers/hwtracing/coresight/coresight-etm4x-core.c
> > >> index 747afc875f91..b5945f62794c 100644
> > >> --- a/drivers/hwtracing/coresight/coresight-etm4x.c
> > >> +++ b/drivers/hwtracing/coresight/coresight-etm4x-core.c
> > >> @@ -1536,6 +1536,26 @@ static struct amba_cs_uci_id uci_id_etm4[] = {
> > >> }
> > >> };
> > >>
> > >> +static int __exit etm4_remove(struct amba_device *adev)
> > >> +{
> > >> + struct etmv4_drvdata *drvdata = dev_get_drvdata(&adev->dev);
> > >> +
> > >> + etm_perf_symlink(drvdata->csdev, false);
> > >> +
> > >> + if (--etm4_count == 0) {
> > >> + etm4_cpu_pm_unregister();
> > >> +
> > >> +
> > > cpuhp_remove_state_nocalls(CPUHP_AP_ARM_CORESIGHT_STARTING);
> > >> + if (hp_online)
> > >> + cpuhp_remove_state_nocalls(hp_online);
> > >> + }
> > >> +
> > >> + coresight_unregister(drvdata->csdev);
> > >> +
> > >> + return 0;
> > >> +}
> > >> +
> > >> +
> > >> static const struct amba_id etm4_ids[] = {
> > >> CS_AMBA_ID(0x000bb95d), /* Cortex-A53 */
> > >> CS_AMBA_ID(0x000bb95e), /* Cortex-A57 */
> > >> @@ -1553,12 +1573,21 @@ static const struct amba_id etm4_ids[] = {
> > >> {},
> > >> };
> > >>
> > >> +MODULE_DEVICE_TABLE(amba, etm4_ids);
> > >> +
> > >> static struct amba_driver etm4x_driver = {
> > >> .drv = {
> > >> .name = "coresight-etm4x",
> > >> + .owner = THIS_MODULE,
> > >> .suppress_bind_attrs = true,
> > >> },
> > >> .probe = etm4_probe,
> > >> + .remove = etm4_remove,
> > >> .id_table = etm4_ids,
> > >> };
> > >> -builtin_amba_driver(etm4x_driver);
> > >> +module_amba_driver(etm4x_driver);
> > >> +
> > >> +MODULE_AUTHOR("Pratik Patel <pratikp at codeaurora.org>");
> > >> +MODULE_AUTHOR("Mathieu Poirier <mathieu.poirier at linaro.org>");
> > >> +MODULE_DESCRIPTION("Arm CoreSight Program Flow Trace v4 driver");
> > >> +MODULE_LICENSE("GPL v2");
> > >> --
> > >> The Qualcomm Innovation Center, Inc. is a member of the Code Aurora
> > > Forum,
> > >> a Linux Foundation Collaborative Project
> > >>
> > >>
> > >> _______________________________________________
> > >> linux-arm-kernel mailing list
> > >> linux-arm-kernel at lists.infradead.org
> > >> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
>
>
>
> --
> Mike Leach
> Principal Engineer, ARM Ltd.
> Manchester Design Centre. UK
More information about the linux-arm-kernel
mailing list