- Linux-stable-mirror - lists.linaro.org

by Greg Kroah-Hartman

This is the start of the stable review cycle for the 4.4.227 release. There are 36 patches in this series, all will be posted as a response to this one. If anyone has any issues with these being applied, please let me know. Responses should be made by Thu, 11 Jun 2020 17:39:17 +0000. Anything received after that time might be too late. The whole patch series can be found in one patch at: https://www.kernel.org/pub/linux/kernel/v4.x/stable-review/patch-4.4.227-rc… or in the git tree and branch at: git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git linux-4.4.y and the diffstat can be found below. thanks, greg k-h ------------- Pseudo-Shortlog of commits: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Linux 4.4.227-rc1 Oleg Nesterov <oleg(a)redhat.com> uprobes: ensure that uprobe->offset and ->ref_ctr_offset are properly aligned Mathieu Othacehe <m.othacehe(a)gmail.com> iio: vcnl4000: Fix i2c swapped word reading. Josh Poimboeuf <jpoimboe(a)redhat.com> x86/speculation: Add Ivy Bridge to affected list Mark Gross <mgross(a)linux.intel.com> x86/speculation: Add SRBDS vulnerability and mitigation documentation Mark Gross <mgross(a)linux.intel.com> x86/speculation: Add Special Register Buffer Data Sampling (SRBDS) mitigation Mark Gross <mgross(a)linux.intel.com> x86/cpu: Add 'table' argument to cpu_matches() Mark Gross <mgross(a)linux.intel.com> x86/cpu: Add a steppings field to struct x86_cpu_id Jia Zhang <qianyue.zj(a)alibaba-inc.com> x86/cpu: Rename cpu_data.x86_mask to cpu_data.x86_stepping Pascal Terjan <pterjan(a)google.com> staging: rtl8712: Fix IEEE80211_ADDBA_PARAM_BUF_SIZE_MASK Dmitry Torokhov <dmitry.torokhov(a)gmail.com> vt: keyboard: avoid signed integer overflow in k_ascii Daniele Palmas <dnlplm(a)gmail.com> USB: serial: option: add Telit LE910C1-EUX compositions Bin Liu <b-liu(a)ti.com> USB: serial: usb_wwan: do not resubmit rx urb on fatal errors Matt Jolly <Kangie(a)footclan.ninja> USB: serial: qcserial: add DW5816e QDL support Eric Dumazet <edumazet(a)google.com> l2tp: do not use inet_hash()/inet_unhash() Eric Dumazet <edumazet(a)google.com> l2tp: add sk_family checks to l2tp_validate_socket Stefano Garzarella <sgarzare(a)redhat.com> vsock: fix timeout in vsock_accept() Chuhong Yuan <hslester96(a)gmail.com> NFC: st21nfca: add missed kfree_skb() in an error path Yang Yingliang <yangyingliang(a)huawei.com> devinet: fix memleak in inetdev_init() Can Guo <cang(a)codeaurora.org> scsi: ufs: Release clock if DMA map fails yangerkun <yangerkun(a)huawei.com> slip: not call free_netdev before rtnl_unlock in slip_open Ben Hutchings <ben(a)decadent.org.uk> slcan: Fix double-free on slcan_open() error path Jérôme Pouiller <jerome.pouiller(a)silabs.com> mmc: fix compilation of user API Guillaume Nault <gnault(a)redhat.com> pppoe: only process PADT targeted at local interfaces Jonathan McDowell <noodles(a)earth.li> net: ethernet: stmmac: Enable interface clocks on probe for IPQ806x Valentin Longchamp <valentin(a)longchamp.me> net/ethernet/freescale: rework quiesce/activate for ucc_geth Jeremy Kerr <jk(a)ozlabs.org> net: bmac: Fix read of MAC address from ROM Nathan Chancellor <natechancellor(a)gmail.com> x86/mmiotrace: Use cpumask_available() for cpumask_var_t variables Eugeniy Paltsev <Eugeniy.Paltsev(a)synopsys.com> ARC: Fix ICCM & DCCM runtime size checks Vasily Gorbik <gor(a)linux.ibm.com> s390/ftrace: save traced function caller Xinwei Kong <kong.kongxinwei(a)hisilicon.com> spi: dw: use "smp_mb()" to avoid sending spi data error Takashi Iwai <tiwai(a)suse.de> ALSA: hda - No loopback on ALC299 codec Christophe Jaillet <christophe.jaillet(a)wanadoo.fr> IB/mlx4: Fix an error handling path in 'mlx4_ib_rereg_user_mr()' Zhen Lei <thunder.leizhen(a)huawei.com> esp6: fix memleak on error path in esp6_input Lee, Chun-Yi <joeyli.kernel(a)gmail.com> platform/x86: acer-wmi: setup accelerometer when ACPI device was found Eugeniu Rosca <erosca(a)de.adit-jv.com> usb: gadget: f_uac2: fix error handling in afunc_bind (again) Hannes Reinecke <hare(a)suse.de> scsi: scsi_devinfo: fixup string compare ------------- Diffstat: Documentation/ABI/testing/sysfs-devices-system-cpu | 1 + .../special-register-buffer-data-sampling.rst | 149 +++++++++++++++++++++ Documentation/kernel-parameters.txt | 20 +++ Makefile | 4 +- arch/arc/kernel/setup.c | 5 +- arch/s390/kernel/mcount.S | 1 + arch/x86/include/asm/acpi.h | 2 +- arch/x86/include/asm/cpu_device_id.h | 27 ++++ arch/x86/include/asm/cpufeatures.h | 2 + arch/x86/include/asm/msr-index.h | 4 + arch/x86/include/asm/processor.h | 2 +- arch/x86/kernel/amd_nb.c | 2 +- arch/x86/kernel/asm-offsets_32.c | 2 +- arch/x86/kernel/cpu/amd.c | 26 ++-- arch/x86/kernel/cpu/bugs.c | 106 +++++++++++++++ arch/x86/kernel/cpu/centaur.c | 4 +- arch/x86/kernel/cpu/common.c | 62 +++++++-- arch/x86/kernel/cpu/cpu.h | 1 + arch/x86/kernel/cpu/cyrix.c | 2 +- arch/x86/kernel/cpu/intel.c | 20 +-- arch/x86/kernel/cpu/match.c | 7 +- arch/x86/kernel/cpu/microcode/intel.c | 4 +- arch/x86/kernel/cpu/mtrr/generic.c | 2 +- arch/x86/kernel/cpu/mtrr/main.c | 4 +- arch/x86/kernel/cpu/perf_event_intel.c | 2 +- arch/x86/kernel/cpu/perf_event_intel_lbr.c | 2 +- arch/x86/kernel/cpu/perf_event_p6.c | 2 +- arch/x86/kernel/cpu/proc.c | 4 +- arch/x86/kernel/head_32.S | 4 +- arch/x86/kernel/mpparse.c | 2 +- arch/x86/mm/mmio-mod.c | 4 +- drivers/base/cpu.c | 8 ++ drivers/char/hw_random/via-rng.c | 2 +- drivers/cpufreq/acpi-cpufreq.c | 2 +- drivers/cpufreq/longhaul.c | 6 +- drivers/cpufreq/p4-clockmod.c | 2 +- drivers/cpufreq/powernow-k7.c | 2 +- drivers/cpufreq/speedstep-centrino.c | 4 +- drivers/cpufreq/speedstep-lib.c | 6 +- drivers/crypto/padlock-aes.c | 2 +- drivers/edac/amd64_edac.c | 2 +- drivers/edac/mce_amd.c | 2 +- drivers/hwmon/coretemp.c | 6 +- drivers/hwmon/hwmon-vid.c | 2 +- drivers/hwmon/k10temp.c | 2 +- drivers/hwmon/k8temp.c | 2 +- drivers/iio/light/vcnl4000.c | 6 +- drivers/infiniband/hw/mlx4/mr.c | 7 +- drivers/net/can/slcan.c | 3 +- drivers/net/ethernet/apple/bmac.c | 2 +- drivers/net/ethernet/freescale/ucc_geth.c | 13 +- .../net/ethernet/stmicro/stmmac/dwmac-ipq806x.c | 13 ++ drivers/net/ppp/pppoe.c | 3 + drivers/net/slip/slip.c | 3 + drivers/nfc/st21nfca/dep.c | 4 +- drivers/platform/x86/acer-wmi.c | 9 +- drivers/scsi/scsi_devinfo.c | 23 ++-- drivers/scsi/ufs/ufshcd.c | 1 + drivers/spi/spi-dw.c | 3 + drivers/staging/rtl8712/wifi.h | 9 +- drivers/tty/vt/keyboard.c | 26 ++-- drivers/usb/gadget/function/f_uac2.c | 4 +- drivers/usb/serial/option.c | 4 + drivers/usb/serial/qcserial.c | 1 + drivers/usb/serial/usb_wwan.c | 4 + drivers/video/fbdev/geode/video_gx.c | 2 +- include/linux/mod_devicetable.h | 6 + include/uapi/linux/mmc/ioctl.h | 1 + kernel/events/uprobes.c | 16 ++- net/ipv4/devinet.c | 1 + net/ipv6/esp6.c | 4 +- net/l2tp/l2tp_core.c | 2 + net/l2tp/l2tp_ip.c | 28 +++- net/l2tp/l2tp_ip6.c | 28 +++- net/vmw_vsock/af_vsock.c | 2 +- sound/pci/hda/patch_realtek.c | 3 + 76 files changed, 603 insertions(+), 157 deletions(-)

5 years, 6 months

1
36
0 0

[PATCH 104/185] drm/i915/gt: Do not schedule normal requests immediately along virtual

by Chris Wilson

When we push a virtual request onto the HW, we update the rq->engine to point to the physical engine. A request that is then submitted by the user that waits upon the virtual engine, but along the physical engine in use, will then see that it is due to be submitted to the same engine and take a shortcut (and be queued without waiting for the completion fence). However, the virtual request may be preempted (either by higher priority users, or by timeslicing) and removed from the physical engine to be migrated over to one of its siblings. The dependent normal request however is oblivious to the removal of the virtual request and remains queued to execute on HW, believing that once it reaches the head of its queue all of its predecessors will have completed executing! v2: Beware restriction of signal->execution_mask prior to submission. Fixes: 6d06779e8672 ("drm/i915: Load balancing across a virtual engine") Testcase: igt/gem_exec_balancer/sliced Signed-off-by: Chris Wilson <chris(a)chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin(a)intel.com> Cc: <stable(a)vger.kernel.org> # v5.3+ Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin(a)intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200526090753.11329-2-chris@… (cherry picked from commit 511b6d9aed417739b6aa49d0b6b4354ad21020f1) --- drivers/gpu/drm/i915/i915_request.c | 25 +++++++++++++++++++++---- 1 file changed, 21 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_request.c b/drivers/gpu/drm/i915/i915_request.c index 406d73bebbcb2..078162ded1a67 100644 --- a/drivers/gpu/drm/i915/i915_request.c +++ b/drivers/gpu/drm/i915/i915_request.c @@ -1242,6 +1242,25 @@ i915_request_await_execution(struct i915_request *rq, return 0; } +static int +await_request_submit(struct i915_request *to, struct i915_request *from) +{ + /* + * If we are waiting on a virtual engine, then it may be + * constrained to execute on a single engine *prior* to submission. + * When it is submitted, it will be first submitted to the virtual + * engine and then passed to the physical engine. We cannot allow + * the waiter to be submitted immediately to the physical engine + * as it may then bypass the virtual request. + */ + if (to->engine == READ_ONCE(from->engine)) + return i915_sw_fence_await_sw_fence_gfp(&to->submit, + &from->submit, + I915_FENCE_GFP); + else + return __i915_request_await_execution(to, from, NULL); +} + static int i915_request_await_request(struct i915_request *to, struct i915_request *from) { @@ -1263,10 +1282,8 @@ i915_request_await_request(struct i915_request *to, struct i915_request *from) return ret; } - if (to->engine == READ_ONCE(from->engine)) - ret = i915_sw_fence_await_sw_fence_gfp(&to->submit, - &from->submit, - I915_FENCE_GFP); + if (is_power_of_2(to->execution_mask | READ_ONCE(from->execution_mask))) + ret = await_request_submit(to, from); else ret = emit_semaphore_wait(to, from, I915_FENCE_GFP); if (ret < 0) --------------------------------------------------------------------- Intel Corporation (UK) Limited Registered No. 1134945 (England) Registered Office: Pipers Way, Swindon SN3 1RJ VAT No: 860 2173 47 This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). Any review or distribution by others is strictly prohibited. If you are not the intended recipient, please contact the sender and delete all copies.

5 years, 6 months

1
0
0 0

[PATCH 087/185] drm/i915: Defer semaphore priority bumping to a workqueue

by Chris Wilson

Since the semaphore fence may be signaled from inside an interrupt handler from inside a request holding its request->lock, we cannot then enter into the engine->active.lock for processing the semaphore priority bump as we may traverse our call tree and end up on another held request. CPU 0: [ 2243.218864] _raw_spin_lock_irqsave+0x9a/0xb0 [ 2243.218867] i915_schedule_bump_priority+0x49/0x80 [i915] [ 2243.218869] semaphore_notify+0x6d/0x98 [i915] [ 2243.218871] __i915_sw_fence_complete+0x61/0x420 [i915] [ 2243.218874] ? kmem_cache_free+0x211/0x290 [ 2243.218876] i915_sw_fence_complete+0x58/0x80 [i915] [ 2243.218879] dma_i915_sw_fence_wake+0x3e/0x80 [i915] [ 2243.218881] signal_irq_work+0x571/0x690 [i915] [ 2243.218883] irq_work_run_list+0xd7/0x120 [ 2243.218885] irq_work_run+0x1d/0x50 [ 2243.218887] smp_irq_work_interrupt+0x21/0x30 [ 2243.218889] irq_work_interrupt+0xf/0x20 CPU 1: [ 2242.173107] _raw_spin_lock+0x8f/0xa0 [ 2242.173110] __i915_request_submit+0x64/0x4a0 [i915] [ 2242.173112] __execlists_submission_tasklet+0x8ee/0x2120 [i915] [ 2242.173114] ? i915_sched_lookup_priolist+0x1e3/0x2b0 [i915] [ 2242.173117] execlists_submit_request+0x2e8/0x2f0 [i915] [ 2242.173119] submit_notify+0x8f/0xc0 [i915] [ 2242.173121] __i915_sw_fence_complete+0x61/0x420 [i915] [ 2242.173124] ? _raw_spin_unlock_irqrestore+0x39/0x40 [ 2242.173137] i915_sw_fence_complete+0x58/0x80 [i915] [ 2242.173140] i915_sw_fence_commit+0x16/0x20 [i915] Closes: https://gitlab.freedesktop.org/drm/intel/issues/1318 Fixes: b7404c7ecb38 ("drm/i915: Bump ready tasks ahead of busywaits") Signed-off-by: Chris Wilson <chris(a)chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin(a)intel.com> Cc: <stable(a)vger.kernel.org> # v5.2+ Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin(a)intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200310101720.9944-1-chris@c… (cherry picked from commit 209df10bb4536c81c2540df96c02cd079435357f) --- drivers/gpu/drm/i915/i915_request.c | 22 +++++++++++++++++----- drivers/gpu/drm/i915/i915_request.h | 2 ++ 2 files changed, 19 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_request.c b/drivers/gpu/drm/i915/i915_request.c index e7f96ce0b071e..c7573e66107d0 100644 --- a/drivers/gpu/drm/i915/i915_request.c +++ b/drivers/gpu/drm/i915/i915_request.c @@ -629,19 +629,31 @@ submit_notify(struct i915_sw_fence *fence, enum i915_sw_fence_notify state) return NOTIFY_DONE; } +static void irq_semaphore_cb(struct irq_work *wrk) +{ + struct i915_request *rq = + container_of(wrk, typeof(*rq), semaphore_work); + + i915_schedule_bump_priority(rq, I915_PRIORITY_NOSEMAPHORE); + i915_request_put(rq); +} + static int __i915_sw_fence_call semaphore_notify(struct i915_sw_fence *fence, enum i915_sw_fence_notify state) { - struct i915_request *request = - container_of(fence, typeof(*request), semaphore); + struct i915_request *rq = container_of(fence, typeof(*rq), semaphore); switch (state) { case FENCE_COMPLETE: - i915_schedule_bump_priority(request, I915_PRIORITY_NOSEMAPHORE); + if (!(READ_ONCE(rq->sched.attr.priority) & I915_PRIORITY_NOSEMAPHORE)) { + i915_request_get(rq); + init_irq_work(&rq->semaphore_work, irq_semaphore_cb); + irq_work_queue(&rq->semaphore_work); + } break; case FENCE_FREE: - i915_request_put(request); + i915_request_put(rq); break; } @@ -1503,9 +1515,9 @@ void __i915_request_queue(struct i915_request *rq, * decide whether to preempt the entire chain so that it is ready to * run at the earliest possible convenience. */ - i915_sw_fence_commit(&rq->semaphore); if (attr && rq->engine->schedule) rq->engine->schedule(rq, attr); + i915_sw_fence_commit(&rq->semaphore); i915_sw_fence_commit(&rq->submit); } diff --git a/drivers/gpu/drm/i915/i915_request.h b/drivers/gpu/drm/i915/i915_request.h index 32ddba894243f..64501b4b3e3a2 100644 --- a/drivers/gpu/drm/i915/i915_request.h +++ b/drivers/gpu/drm/i915/i915_request.h @@ -26,6 +26,7 @@ #define I915_REQUEST_H #include <linux/dma-fence.h> +#include <linux/irq_work.h> #include <linux/lockdep.h> #include "gem/i915_gem_context_types.h" @@ -208,6 +209,7 @@ struct i915_request { }; struct list_head execute_cb; struct i915_sw_fence semaphore; + struct irq_work semaphore_work; /* * A list of everyone we wait upon, and everyone who waits upon us. --------------------------------------------------------------------- Intel Corporation (UK) Limited Registered No. 1134945 (England) Registered Office: Pipers Way, Swindon SN3 1RJ VAT No: 860 2173 47 This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). Any review or distribution by others is strictly prohibited. If you are not the intended recipient, please contact the sender and delete all copies.

5 years, 6 months

1
0
0 0

[PATCH 076/185] drm/i915: Mark concurrent submissions with a weak-dependency

by Chris Wilson

We recorded the dependencies for WAIT_FOR_SUBMIT in order that we could correctly perform priority inheritance from the parallel branches to the common trunk. However, for the purpose of timeslicing and reset handling, the dependency is weak -- as we the pair of requests are allowed to run in parallel and not in strict succession. The real significance though is that this allows us to rearrange groups of WAIT_FOR_SUBMIT linked requests along the single engine, and so can resolve user level inter-batch scheduling dependencies from user semaphores. Fixes: c81471f5e95c ("drm/i915: Copy across scheduler behaviour flags across submit fences") Testcase: igt/gem_exec_fence/submit Signed-off-by: Chris Wilson <chris(a)chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin(a)intel.com> Cc: <stable(a)vger.kernel.org> # v5.6+ Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin(a)intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200507155109.8892-1-chris@c… (cherry picked from commit 6b6cd2ebd8d071e55998e32b648bb8081f7f02bb) --- drivers/gpu/drm/i915/gt/intel_lrc.c | 3 +++ drivers/gpu/drm/i915/i915_request.c | 8 ++++++-- drivers/gpu/drm/i915/i915_scheduler.c | 6 +++--- drivers/gpu/drm/i915/i915_scheduler.h | 3 ++- drivers/gpu/drm/i915/i915_scheduler_types.h | 1 + 5 files changed, 15 insertions(+), 6 deletions(-) diff --git a/drivers/gpu/drm/i915/gt/intel_lrc.c b/drivers/gpu/drm/i915/gt/intel_lrc.c index 2237d03545440..e57b827b00b1e 100644 --- a/drivers/gpu/drm/i915/gt/intel_lrc.c +++ b/drivers/gpu/drm/i915/gt/intel_lrc.c @@ -1902,6 +1902,9 @@ static void defer_request(struct i915_request *rq, struct list_head * const pl) struct i915_request *w = container_of(p->waiter, typeof(*w), sched); + if (p->flags & I915_DEPENDENCY_WEAK) + continue; + /* Leave semaphores spinning on the other engines */ if (w->engine != rq->engine) continue; diff --git a/drivers/gpu/drm/i915/i915_request.c b/drivers/gpu/drm/i915/i915_request.c index ee406fa6774ad..ed1af57d1db18 100644 --- a/drivers/gpu/drm/i915/i915_request.c +++ b/drivers/gpu/drm/i915/i915_request.c @@ -1048,7 +1048,9 @@ i915_request_await_request(struct i915_request *to, struct i915_request *from) return 0; if (to->engine->schedule) { - ret = i915_sched_node_add_dependency(&to->sched, &from->sched); + ret = i915_sched_node_add_dependency(&to->sched, + &from->sched, + I915_DEPENDENCY_EXTERNAL); if (ret < 0) return ret; } @@ -1179,7 +1181,9 @@ __i915_request_await_execution(struct i915_request *to, /* Couple the dependency tree for PI on this exposed to->fence */ if (to->engine->schedule) { - err = i915_sched_node_add_dependency(&to->sched, &from->sched); + err = i915_sched_node_add_dependency(&to->sched, + &from->sched, + I915_DEPENDENCY_WEAK); if (err < 0) return err; } diff --git a/drivers/gpu/drm/i915/i915_scheduler.c b/drivers/gpu/drm/i915/i915_scheduler.c index e99423e548c99..09d6d4538ff38 100644 --- a/drivers/gpu/drm/i915/i915_scheduler.c +++ b/drivers/gpu/drm/i915/i915_scheduler.c @@ -463,7 +463,8 @@ bool __i915_sched_node_add_dependency(struct i915_sched_node *node, } int i915_sched_node_add_dependency(struct i915_sched_node *node, - struct i915_sched_node *signal) + struct i915_sched_node *signal, + unsigned long flags) { struct i915_dependency *dep; @@ -472,8 +473,7 @@ int i915_sched_node_add_dependency(struct i915_sched_node *node, return -ENOMEM; if (!__i915_sched_node_add_dependency(node, signal, dep, - I915_DEPENDENCY_EXTERNAL | - I915_DEPENDENCY_ALLOC)) + flags | I915_DEPENDENCY_ALLOC)) i915_dependency_free(dep); return 0; diff --git a/drivers/gpu/drm/i915/i915_scheduler.h b/drivers/gpu/drm/i915/i915_scheduler.h index d1dc4efef77b5..6f0bf00fc5690 100644 --- a/drivers/gpu/drm/i915/i915_scheduler.h +++ b/drivers/gpu/drm/i915/i915_scheduler.h @@ -34,7 +34,8 @@ bool __i915_sched_node_add_dependency(struct i915_sched_node *node, unsigned long flags); int i915_sched_node_add_dependency(struct i915_sched_node *node, - struct i915_sched_node *signal); + struct i915_sched_node *signal, + unsigned long flags); void i915_sched_node_fini(struct i915_sched_node *node); diff --git a/drivers/gpu/drm/i915/i915_scheduler_types.h b/drivers/gpu/drm/i915/i915_scheduler_types.h index d18e705500542..7186875088a0a 100644 --- a/drivers/gpu/drm/i915/i915_scheduler_types.h +++ b/drivers/gpu/drm/i915/i915_scheduler_types.h @@ -78,6 +78,7 @@ struct i915_dependency { unsigned long flags; #define I915_DEPENDENCY_ALLOC BIT(0) #define I915_DEPENDENCY_EXTERNAL BIT(1) +#define I915_DEPENDENCY_WEAK BIT(2) }; #endif /* _I915_SCHEDULER_TYPES_H_ */ --------------------------------------------------------------------- Intel Corporation (UK) Limited Registered No. 1134945 (England) Registered Office: Pipers Way, Swindon SN3 1RJ VAT No: 860 2173 47 This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). Any review or distribution by others is strictly prohibited. If you are not the intended recipient, please contact the sender and delete all copies.

5 years, 6 months

1
0
0 0

[PATCH 060/185] drm/i915/execlists: Track inflight CCID

by Chris Wilson

The presumption is that by using a circular counter that is twice as large as the maximum ELSP submission, we would never reuse the same CCID for two inflight contexts. However, if we continually preempt an active context such that it always remains inflight, it can be resubmitted with an arbitrary number of paired contexts. As each of its paired contexts will use a new CCID, eventually it will wrap and submit two ELSP with the same CCID. Rather than use a simple circular counter, switch over to a small bitmap of inflight ids so we can avoid reusing one that is still potentially active. Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/1796 Fixes: 2935ed5339c4 ("drm/i915: Remove logical HW ID") Signed-off-by: Chris Wilson <chris(a)chris-wilson.co.uk> Cc: Mika Kuoppala <mika.kuoppala(a)linux.intel.com> Cc: <stable(a)vger.kernel.org> # v5.5+ Reviewed-by: Mika Kuoppala <mika.kuoppala(a)linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200428184751.11257-2-chris@… (cherry picked from commit 5c4a53e3b1cbc38d0906e382f1037290658759bb) --- drivers/gpu/drm/i915/gt/intel_engine_types.h | 3 +- drivers/gpu/drm/i915/gt/intel_lrc.c | 29 +++++++++++++++----- drivers/gpu/drm/i915/i915_perf.c | 3 +- drivers/gpu/drm/i915/selftests/i915_vma.c | 2 +- 4 files changed, 25 insertions(+), 12 deletions(-) diff --git a/drivers/gpu/drm/i915/gt/intel_engine_types.h b/drivers/gpu/drm/i915/gt/intel_engine_types.h index e0de4af9a3481..781f94ce65f65 100644 --- a/drivers/gpu/drm/i915/gt/intel_engine_types.h +++ b/drivers/gpu/drm/i915/gt/intel_engine_types.h @@ -308,8 +308,7 @@ struct intel_engine_cs { u32 context_size; u32 mmio_base; - unsigned int context_tag; -#define NUM_CONTEXT_TAG roundup_pow_of_two(2 * EXECLIST_MAX_PORTS) + unsigned long context_tag; struct rb_node uabi_node; diff --git a/drivers/gpu/drm/i915/gt/intel_lrc.c b/drivers/gpu/drm/i915/gt/intel_lrc.c index dfcc44171df1c..adfaf52ca8d08 100644 --- a/drivers/gpu/drm/i915/gt/intel_lrc.c +++ b/drivers/gpu/drm/i915/gt/intel_lrc.c @@ -1321,13 +1321,17 @@ __execlists_schedule_in(struct i915_request *rq) if (ce->tag) { /* Use a fixed tag for OA and friends */ + GEM_BUG_ON(ce->tag <= BITS_PER_LONG); ce->lrc.ccid = ce->tag; } else { /* We don't need a strict matching tag, just different values */ - ce->lrc.ccid = - (++engine->context_tag % NUM_CONTEXT_TAG) << - (GEN11_SW_CTX_ID_SHIFT - 32); - BUILD_BUG_ON(NUM_CONTEXT_TAG > GEN12_MAX_CONTEXT_HW_ID); + unsigned int tag = ffs(engine->context_tag); + + GEM_BUG_ON(tag == 0 || tag >= BITS_PER_LONG); + clear_bit(tag - 1, &engine->context_tag); + ce->lrc.ccid = tag << (GEN11_SW_CTX_ID_SHIFT - 32); + + BUILD_BUG_ON(BITS_PER_LONG > GEN12_MAX_CONTEXT_HW_ID); } ce->lrc.ccid |= engine->execlists.ccid; @@ -1371,7 +1375,8 @@ static void kick_siblings(struct i915_request *rq, struct intel_context *ce) static inline void __execlists_schedule_out(struct i915_request *rq, - struct intel_engine_cs * const engine) + struct intel_engine_cs * const engine, + unsigned int ccid) { struct intel_context * const ce = rq->context; @@ -1389,6 +1394,14 @@ __execlists_schedule_out(struct i915_request *rq, i915_request_completed(rq)) intel_engine_add_retire(engine, ce->timeline); + ccid >>= GEN11_SW_CTX_ID_SHIFT - 32; + ccid &= GEN12_MAX_CONTEXT_HW_ID; + if (ccid < BITS_PER_LONG) { + GEM_BUG_ON(ccid == 0); + GEM_BUG_ON(test_bit(ccid - 1, &engine->context_tag)); + set_bit(ccid - 1, &engine->context_tag); + } + intel_context_update_runtime(ce); intel_engine_context_out(engine); execlists_context_status_change(rq, INTEL_CONTEXT_SCHEDULE_OUT); @@ -1414,15 +1427,17 @@ execlists_schedule_out(struct i915_request *rq) { struct intel_context * const ce = rq->context; struct intel_engine_cs *cur, *old; + u32 ccid; trace_i915_request_out(rq); + ccid = rq->context->lrc.ccid; old = READ_ONCE(ce->inflight); do cur = ptr_unmask_bits(old, 2) ? ptr_dec(old) : NULL; while (!try_cmpxchg(&ce->inflight, &old, cur)); if (!cur) - __execlists_schedule_out(rq, old); + __execlists_schedule_out(rq, old, ccid); i915_request_put(rq); } @@ -3883,7 +3898,7 @@ static void enable_execlists(struct intel_engine_cs *engine) enable_error_interrupt(engine); - engine->context_tag = 0; + engine->context_tag = GENMASK(BITS_PER_LONG - 2, 0); } static bool unexpected_starting_state(struct intel_engine_cs *engine) diff --git a/drivers/gpu/drm/i915/i915_perf.c b/drivers/gpu/drm/i915/i915_perf.c index 6e3d378ebbbd6..820c6ba755cce 100644 --- a/drivers/gpu/drm/i915/i915_perf.c +++ b/drivers/gpu/drm/i915/i915_perf.c @@ -1326,11 +1326,10 @@ static int oa_get_render_ctx_id(struct i915_perf_stream *stream) ((1U << GEN11_SW_CTX_ID_WIDTH) - 1) << (GEN11_SW_CTX_ID_SHIFT - 32); /* * Pick an unused context id - * 0 - (NUM_CONTEXT_TAG - 1) are used by other contexts + * 0 - BITS_PER_LONG are used by other contexts * GEN12_MAX_CONTEXT_HW_ID (0x7ff) is used by idle context */ stream->specific_ctx_id = (GEN12_MAX_CONTEXT_HW_ID - 1) << (GEN11_SW_CTX_ID_SHIFT - 32); - BUILD_BUG_ON((GEN12_MAX_CONTEXT_HW_ID - 1) < NUM_CONTEXT_TAG); break; } diff --git a/drivers/gpu/drm/i915/selftests/i915_vma.c b/drivers/gpu/drm/i915/selftests/i915_vma.c index 58b5f40a07dd6..af89c7fc8f593 100644 --- a/drivers/gpu/drm/i915/selftests/i915_vma.c +++ b/drivers/gpu/drm/i915/selftests/i915_vma.c @@ -173,7 +173,7 @@ static int igt_vma_create(void *arg) } nc = 0; - for_each_prime_number(num_ctx, 2 * NUM_CONTEXT_TAG) { + for_each_prime_number(num_ctx, 2 * BITS_PER_LONG) { for (; nc < num_ctx; nc++) { ctx = mock_context(i915, "mock"); if (!ctx) --------------------------------------------------------------------- Intel Corporation (UK) Limited Registered No. 1134945 (England) Registered Office: Pipers Way, Swindon SN3 1RJ VAT No: 860 2173 47 This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). Any review or distribution by others is strictly prohibited. If you are not the intended recipient, please contact the sender and delete all copies.

5 years, 6 months

1
0
0 0

[PATCH 059/185] drm/i915/execlists: Avoid reusing the same logical CCID

by Chris Wilson

The bspec is confusing on the nature of the upper 32bits of the LRC descriptor. Once upon a time, it said that it uses the upper 32b to decide if it should perform a lite-restore, and so we must ensure that each unique context submitted to HW is given a unique CCID [for the duration of it being on the HW]. Currently, this is achieved by using a small circular tag, and assigning every context submitted to HW a new id. However, this tag is being cleared on repinning an inflight context such that we end up re-using the 0 tag for multiple contexts. To avoid accidentally clearing the CCID in the upper 32bits of the LRC descriptor, split the descriptor into two dwords so we can update the GGTT address separately from the CCID. Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/1796 Fixes: 2935ed5339c4 ("drm/i915: Remove logical HW ID") Signed-off-by: Chris Wilson <chris(a)chris-wilson.co.uk> Cc: Mika Kuoppala <mika.kuoppala(a)linux.intel.com> Cc: <stable(a)vger.kernel.org> # v5.5+ Reviewed-by: Mika Kuoppala <mika.kuoppala(a)linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200428184751.11257-1-chris@… (cherry picked from commit 2632f174a2e1a5fd40a70404fa8ccfd0b1f79ebd) --- drivers/gpu/drm/i915/gt/intel_context_types.h | 8 ++- drivers/gpu/drm/i915/gt/intel_engine_cs.c | 6 +- drivers/gpu/drm/i915/gt/intel_engine_types.h | 5 ++ drivers/gpu/drm/i915/gt/intel_lrc.c | 57 ++++++++----------- .../gpu/drm/i915/gt/uc/intel_guc_submission.c | 2 +- drivers/gpu/drm/i915/gvt/scheduler.c | 4 +- drivers/gpu/drm/i915/i915_perf.c | 3 +- 7 files changed, 44 insertions(+), 41 deletions(-) diff --git a/drivers/gpu/drm/i915/gt/intel_context_types.h b/drivers/gpu/drm/i915/gt/intel_context_types.h index e887a2ff28c50..45d257be44308 100644 --- a/drivers/gpu/drm/i915/gt/intel_context_types.h +++ b/drivers/gpu/drm/i915/gt/intel_context_types.h @@ -69,7 +69,13 @@ struct intel_context { #define CONTEXT_NOPREEMPT 7 u32 *lrc_reg_state; - u64 lrc_desc; + union { + struct { + u32 lrca; + u32 ccid; + }; + u64 desc; + } lrc; u32 tag; /* cookie passed to HW to track this context on submission */ /* Time on GPU as tracked by the hw. */ diff --git a/drivers/gpu/drm/i915/gt/intel_engine_cs.c b/drivers/gpu/drm/i915/gt/intel_engine_cs.c index 1a2887fac48d6..a1bd3e3302774 100644 --- a/drivers/gpu/drm/i915/gt/intel_engine_cs.c +++ b/drivers/gpu/drm/i915/gt/intel_engine_cs.c @@ -1411,9 +1411,9 @@ static void intel_engine_print_registers(struct intel_engine_cs *engine, int len; len = scnprintf(hdr, sizeof(hdr), - "\t\tActive[%d]: ccid:%08x, ", + "\t\tActive[%d]: ccid:%08x, ", (int)(port - execlists->active), - upper_32_bits(rq->context->lrc_desc)); + rq->context->lrc.ccid); len += print_ring(hdr + len, sizeof(hdr) - len, rq); scnprintf(hdr + len, sizeof(hdr) - len, "rq: "); print_request(m, rq, hdr); @@ -1425,7 +1425,7 @@ static void intel_engine_print_registers(struct intel_engine_cs *engine, len = scnprintf(hdr, sizeof(hdr), "\t\tPending[%d]: ccid:%08x, ", (int)(port - execlists->pending), - upper_32_bits(rq->context->lrc_desc)); + rq->context->lrc.ccid); len += print_ring(hdr + len, sizeof(hdr) - len, rq); scnprintf(hdr + len, sizeof(hdr) - len, "rq: "); print_request(m, rq, hdr); diff --git a/drivers/gpu/drm/i915/gt/intel_engine_types.h b/drivers/gpu/drm/i915/gt/intel_engine_types.h index 26ba9f6ab0c7a..e0de4af9a3481 100644 --- a/drivers/gpu/drm/i915/gt/intel_engine_types.h +++ b/drivers/gpu/drm/i915/gt/intel_engine_types.h @@ -155,6 +155,11 @@ struct intel_engine_execlists { */ struct i915_priolist default_priolist; + /** + * @ccid: identifier for contexts submitted to this engine + */ + u32 ccid; + /** * @yield: CCID at the time of the last semaphore-wait interrupt. * diff --git a/drivers/gpu/drm/i915/gt/intel_lrc.c b/drivers/gpu/drm/i915/gt/intel_lrc.c index 59569bc5b6a48..dfcc44171df1c 100644 --- a/drivers/gpu/drm/i915/gt/intel_lrc.c +++ b/drivers/gpu/drm/i915/gt/intel_lrc.c @@ -560,10 +560,10 @@ assert_priority_queue(const struct i915_request *prev, * engine info, SW context ID and SW counter need to form a unique number * (Context ID) per lrc. */ -static u64 +static u32 lrc_descriptor(struct intel_context *ce, struct intel_engine_cs *engine) { - u64 desc; + u32 desc; desc = INTEL_LEGACY_32B_CONTEXT; if (i915_vm_is_4lvl(ce->vm)) @@ -574,21 +574,7 @@ lrc_descriptor(struct intel_context *ce, struct intel_engine_cs *engine) if (IS_GEN(engine->i915, 8)) desc |= GEN8_CTX_L3LLC_COHERENT; - desc |= i915_ggtt_offset(ce->state); /* bits 12-31 */ - /* - * The following 32bits are copied into the OA reports (dword 2). - * Consider updating oa_get_render_ctx_id in i915_perf.c when changing - * anything below. - */ - if (INTEL_GEN(engine->i915) >= 11) { - desc |= (u64)engine->instance << GEN11_ENGINE_INSTANCE_SHIFT; - /* bits 48-53 */ - - desc |= (u64)engine->class << GEN11_ENGINE_CLASS_SHIFT; - /* bits 61-63 */ - } - - return desc; + return i915_ggtt_offset(ce->state) | desc; } static inline unsigned int dword_in_page(void *addr) @@ -1285,7 +1271,7 @@ static void reset_active(struct i915_request *rq, __execlists_update_reg_state(ce, engine, head); /* We've switched away, so this should be a no-op, but intent matters */ - ce->lrc_desc |= CTX_DESC_FORCE_RESTORE; + ce->lrc.desc |= CTX_DESC_FORCE_RESTORE; } static void st_update_runtime_underflow(struct intel_context *ce, s32 dt) @@ -1333,18 +1319,19 @@ __execlists_schedule_in(struct i915_request *rq) if (IS_ENABLED(CONFIG_DRM_I915_DEBUG_GEM)) execlists_check_context(ce, engine); - ce->lrc_desc &= ~GENMASK_ULL(47, 37); if (ce->tag) { /* Use a fixed tag for OA and friends */ - ce->lrc_desc |= (u64)ce->tag << 32; + ce->lrc.ccid = ce->tag; } else { /* We don't need a strict matching tag, just different values */ - ce->lrc_desc |= - (u64)(++engine->context_tag % NUM_CONTEXT_TAG) << - GEN11_SW_CTX_ID_SHIFT; + ce->lrc.ccid = + (++engine->context_tag % NUM_CONTEXT_TAG) << + (GEN11_SW_CTX_ID_SHIFT - 32); BUILD_BUG_ON(NUM_CONTEXT_TAG > GEN12_MAX_CONTEXT_HW_ID); } + ce->lrc.ccid |= engine->execlists.ccid; + __intel_gt_pm_get(engine->gt); execlists_context_status_change(rq, INTEL_CONTEXT_SCHEDULE_IN); intel_engine_context_in(engine); @@ -1443,7 +1430,7 @@ execlists_schedule_out(struct i915_request *rq) static u64 execlists_update_context(struct i915_request *rq) { struct intel_context *ce = rq->context; - u64 desc = ce->lrc_desc; + u64 desc = ce->lrc.desc; u32 tail, prev; /* @@ -1482,7 +1469,7 @@ static u64 execlists_update_context(struct i915_request *rq) */ wmb(); - ce->lrc_desc &= ~CTX_DESC_FORCE_RESTORE; + ce->lrc.desc &= ~CTX_DESC_FORCE_RESTORE; return desc; } @@ -1503,8 +1490,9 @@ dump_port(char *buf, int buflen, const char *prefix, struct i915_request *rq) if (!rq) return ""; - snprintf(buf, buflen, "%s%llx:%lld%s prio %d", + snprintf(buf, buflen, "%sccid:%x %llx:%lld%s prio %d", prefix, + rq->context->lrc.ccid, rq->fence.context, rq->fence.seqno, i915_request_completed(rq) ? "!" : i915_request_started(rq) ? "*" : @@ -1880,7 +1868,7 @@ timeslice_yield(const struct intel_engine_execlists *el, * safe, yield if it might be stuck -- it will be given a fresh * timeslice in the near future. */ - return upper_32_bits(rq->context->lrc_desc) == READ_ONCE(el->yield); + return rq->context->lrc.ccid == READ_ONCE(el->yield); } static bool @@ -2907,7 +2895,7 @@ active_context(struct intel_engine_cs *engine, u32 ccid) */ for (port = el->active; (rq = *port); port++) { - if (upper_32_bits(rq->context->lrc_desc) == ccid) { + if (rq->context->lrc.ccid == ccid) { ENGINE_TRACE(engine, "ccid found at active:%zd\n", port - el->active); @@ -2916,7 +2904,7 @@ active_context(struct intel_engine_cs *engine, u32 ccid) } for (port = el->pending; (rq = *port); port++) { - if (upper_32_bits(rq->context->lrc_desc) == ccid) { + if (rq->context->lrc.ccid == ccid) { ENGINE_TRACE(engine, "ccid found at pending:%zd\n", port - el->pending); @@ -3337,7 +3325,7 @@ __execlists_context_pin(struct intel_context *ce, if (IS_ERR(vaddr)) return PTR_ERR(vaddr); - ce->lrc_desc = lrc_descriptor(ce, engine) | CTX_DESC_FORCE_RESTORE; + ce->lrc.lrca = lrc_descriptor(ce, engine) | CTX_DESC_FORCE_RESTORE; ce->lrc_reg_state = vaddr + LRC_STATE_OFFSET; __execlists_update_reg_state(ce, engine, ce->ring->tail); @@ -3366,7 +3354,7 @@ static void execlists_context_reset(struct intel_context *ce) ce, ce->engine, ce->ring, true); __execlists_update_reg_state(ce, ce->engine, ce->ring->tail); - ce->lrc_desc |= CTX_DESC_FORCE_RESTORE; + ce->lrc.desc |= CTX_DESC_FORCE_RESTORE; } static const struct intel_context_ops execlists_context_ops = { @@ -4077,7 +4065,7 @@ static void __execlists_reset(struct intel_engine_cs *engine, bool stalled) head, ce->ring->tail); __execlists_reset_reg_state(ce, engine); __execlists_update_reg_state(ce, engine, head); - ce->lrc_desc |= CTX_DESC_FORCE_RESTORE; /* paranoid: GPU was reset! */ + ce->lrc.desc |= CTX_DESC_FORCE_RESTORE; /* paranoid: GPU was reset! */ unwind: /* Push back any incomplete requests for replay after the reset. */ @@ -4880,6 +4868,11 @@ int intel_execlists_submission_setup(struct intel_engine_cs *engine) else execlists->csb_size = GEN11_CSB_ENTRIES; + if (INTEL_GEN(engine->i915) >= 11) { + execlists->ccid |= engine->instance << (GEN11_ENGINE_INSTANCE_SHIFT - 32); + execlists->ccid |= engine->class << (GEN11_ENGINE_CLASS_SHIFT - 32); + } + /* Finally, take ownership and responsibility for cleanup! */ engine->sanitize = execlists_sanitize; engine->release = execlists_release; diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c b/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c index fe7778c28d2d7..aa6d56e25a10a 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c +++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c @@ -217,7 +217,7 @@ static void guc_wq_item_append(struct intel_guc *guc, static void guc_add_request(struct intel_guc *guc, struct i915_request *rq) { struct intel_engine_cs *engine = rq->engine; - u32 ctx_desc = lower_32_bits(rq->context->lrc_desc); + u32 ctx_desc = rq->context->lrc.ccid; u32 ring_tail = intel_ring_set_tail(rq->ring, rq->tail) / sizeof(u64); guc_wq_item_append(guc, engine->guc_id, ctx_desc, diff --git a/drivers/gpu/drm/i915/gvt/scheduler.c b/drivers/gpu/drm/i915/gvt/scheduler.c index 68932b33c0577..9be89b8d952c1 100644 --- a/drivers/gpu/drm/i915/gvt/scheduler.c +++ b/drivers/gpu/drm/i915/gvt/scheduler.c @@ -286,7 +286,7 @@ static void shadow_context_descriptor_update(struct intel_context *ce, struct intel_vgpu_workload *workload) { - u64 desc = ce->lrc_desc; + u64 desc = ce->lrc.desc; /* * Update bits 0-11 of the context descriptor which includes flags @@ -296,7 +296,7 @@ shadow_context_descriptor_update(struct intel_context *ce, desc |= workload->ctx_desc.addressing_mode << GEN8_CTX_ADDRESSING_MODE_SHIFT; - ce->lrc_desc = desc; + ce->lrc.desc = desc; } static int copy_workload_to_ring_buffer(struct intel_vgpu_workload *workload) diff --git a/drivers/gpu/drm/i915/i915_perf.c b/drivers/gpu/drm/i915/i915_perf.c index a5bddcc5bb334..6e3d378ebbbd6 100644 --- a/drivers/gpu/drm/i915/i915_perf.c +++ b/drivers/gpu/drm/i915/i915_perf.c @@ -1309,8 +1309,7 @@ static int oa_get_render_ctx_id(struct i915_perf_stream *stream) * dropped by GuC. They won't be part of the context * ID in the OA reports, so squash those lower bits. */ - stream->specific_ctx_id = - lower_32_bits(ce->lrc_desc) >> 12; + stream->specific_ctx_id = ce->lrc.lrca >> 12; /* * GuC uses the top bit to signal proxy submission, so --------------------------------------------------------------------- Intel Corporation (UK) Limited Registered No. 1134945 (England) Registered Office: Pipers Way, Swindon SN3 1RJ VAT No: 860 2173 47 This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). Any review or distribution by others is strictly prohibited. If you are not the intended recipient, please contact the sender and delete all copies.

5 years, 6 months

1
0
0 0

[PATCH 030/185] drm/i915/execlists: Enable timeslice on partial virtual engine dequeue

by Chris Wilson

If we stop filling the ELSP due to an incompatible virtual engine request, check if we should enable the timeslice on behalf of the queue. This fixes the case where we are inspecting the last->next element when we know that the last element is the last request in the execution queue, and so decided we did not need to enable timeslicing despite the intent to do so! Fixes: 8ee36e048c98 ("drm/i915/execlists: Minimalistic timeslicing") Signed-off-by: Chris Wilson <chris(a)chris-wilson.co.uk> Cc: Mika Kuoppala <mika.kuoppala(a)linux.intel.com> Cc: Tvrtko Ursulin <tvrtko.ursulin(a)intel.com> Cc: <stable(a)vger.kernel.org> # v5.4+ Reviewed-by: Mika Kuoppala <mika.kuoppala(a)linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200306113012.3184606-1-chri… (cherry picked from commit 3df2deed411e0f1b7312baf0139aab8bba4c0410) --- drivers/gpu/drm/i915/gt/intel_lrc.c | 29 ++++++++++++++++++----------- 1 file changed, 18 insertions(+), 11 deletions(-) diff --git a/drivers/gpu/drm/i915/gt/intel_lrc.c b/drivers/gpu/drm/i915/gt/intel_lrc.c index 354c68b74cec0..114c2b070a6d9 100644 --- a/drivers/gpu/drm/i915/gt/intel_lrc.c +++ b/drivers/gpu/drm/i915/gt/intel_lrc.c @@ -1839,11 +1839,9 @@ need_timeslice(struct intel_engine_cs *engine, const struct i915_request *rq) if (!intel_engine_has_timeslices(engine)) return false; - if (list_is_last(&rq->sched.link, &engine->active.requests)) - return false; - - hint = max(rq_prio(list_next_entry(rq, sched.link)), - engine->execlists.queue_priority_hint); + hint = engine->execlists.queue_priority_hint; + if (!list_is_last(&rq->sched.link, &engine->active.requests)) + hint = max(hint, rq_prio(list_next_entry(rq, sched.link))); return hint >= effective_prio(rq); } @@ -1885,6 +1883,18 @@ static void set_timeslice(struct intel_engine_cs *engine) set_timer_ms(&engine->execlists.timer, active_timeslice(engine)); } +static void start_timeslice(struct intel_engine_cs *engine) +{ + struct intel_engine_execlists *execlists = &engine->execlists; + + execlists->switch_priority_hint = execlists->queue_priority_hint; + + if (timer_pending(&execlists->timer)) + return; + + set_timer_ms(&execlists->timer, timeslice(engine)); +} + static void record_preemption(struct intel_engine_execlists *execlists) { (void)I915_SELFTEST_ONLY(execlists->preempt_hang.count++); @@ -2048,11 +2058,7 @@ static void execlists_dequeue(struct intel_engine_cs *engine) * Even if ELSP[1] is occupied and not worthy * of timeslices, our queue might be. */ - if (!execlists->timer.expires && - need_timeslice(engine, last)) - set_timer_ms(&execlists->timer, - timeslice(engine)); - + start_timeslice(engine); return; } } @@ -2087,7 +2093,8 @@ static void execlists_dequeue(struct intel_engine_cs *engine) if (last && !can_merge_rq(last, rq)) { spin_unlock(&ve->base.active.lock); - return; /* leave this for another */ + start_timeslice(engine); + return; /* leave this for another sibling */ } ENGINE_TRACE(engine, --------------------------------------------------------------------- Intel Corporation (UK) Limited Registered No. 1134945 (England) Registered Office: Pipers Way, Swindon SN3 1RJ VAT No: 860 2173 47 This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). Any review or distribution by others is strictly prohibited. If you are not the intended recipient, please contact the sender and delete all copies.

5 years, 6 months

1
0
0 0

[PATCH 025/185] drm/i915: Return early for await_start on same timeline

by Chris Wilson

Requests within a timeline are ordered by that timeline, so awaiting for the start of a request within the timeline is a no-op. This used to work by falling out of the mutex_trylock() as the signaler and waiter had the same timeline and not returning an error. Fixes: 6a79d848403d ("drm/i915: Lock signaler timeline while navigating") Signed-off-by: Chris Wilson <chris(a)chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin(a)linux.intel.com> Cc: <stable(a)vger.kernel.org> # v5.5+ Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin(a)intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200305134822.2750496-1-chri… (cherry picked from commit ab7a69020fb5d5c7ba19fba60f62fd6f9ca9f779) --- drivers/gpu/drm/i915/i915_request.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_request.c b/drivers/gpu/drm/i915/i915_request.c index 6c387fa1a8547..1213009f31fbe 100644 --- a/drivers/gpu/drm/i915/i915_request.c +++ b/drivers/gpu/drm/i915/i915_request.c @@ -830,8 +830,8 @@ i915_request_await_start(struct i915_request *rq, struct i915_request *signal) struct dma_fence *fence; int err; - GEM_BUG_ON(i915_request_timeline(rq) == - rcu_access_pointer(signal->timeline)); + if (i915_request_timeline(rq) == rcu_access_pointer(signal->timeline)) + return 0; if (i915_request_started(signal)) return 0; --------------------------------------------------------------------- Intel Corporation (UK) Limited Registered No. 1134945 (England) Registered Office: Pipers Way, Swindon SN3 1RJ VAT No: 860 2173 47 This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). Any review or distribution by others is strictly prohibited. If you are not the intended recipient, please contact the sender and delete all copies.

5 years, 6 months

1
0
0 0

[PATCH 024/185] drm/i915: Actually emit the await_start

by Chris Wilson

Fix the inverted test to emit the wait on the end of the previous request if we /haven't/ already. Fixes: 6a79d848403d ("drm/i915: Lock signaler timeline while navigating") Signed-off-by: Chris Wilson <chris(a)chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin(a)linux.intel.com> Cc: <stable(a)vger.kernel.org> # v5.5+ Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin(a)intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200305104210.2619967-1-chri… (cherry picked from commit 07e9c59d63df6a1c44c1975c01827ba18b69270a) --- drivers/gpu/drm/i915/i915_request.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/i915_request.c b/drivers/gpu/drm/i915/i915_request.c index 191a538afa5a8..6c387fa1a8547 100644 --- a/drivers/gpu/drm/i915/i915_request.c +++ b/drivers/gpu/drm/i915/i915_request.c @@ -875,7 +875,7 @@ i915_request_await_start(struct i915_request *rq, struct i915_request *signal) return 0; err = 0; - if (intel_timeline_sync_is_later(i915_request_timeline(rq), fence)) + if (!intel_timeline_sync_is_later(i915_request_timeline(rq), fence)) err = i915_sw_fence_await_dma_fence(&rq->submit, fence, 0, I915_FENCE_GFP); --------------------------------------------------------------------- Intel Corporation (UK) Limited Registered No. 1134945 (England) Registered Office: Pipers Way, Swindon SN3 1RJ VAT No: 860 2173 47 This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). Any review or distribution by others is strictly prohibited. If you are not the intended recipient, please contact the sender and delete all copies.

5 years, 6 months

1
0
0 0

[PATCH 011/185] drm/i915/execlists: Always force a context reload when rewinding RING_TAIL

by Chris Wilson

If we rewind the RING_TAIL on a context, due to a preemption event, we must force the context restore for the RING_TAIL update to be properly handled. Rather than note which preemption events may cause us to rewind the tail, compare the new request's tail with the previously submitted RING_TAIL, as it turns out that timeslicing was causing unexpected rewinds. <idle>-0 0d.s2 1280851190us : __execlists_submission_tasklet: 0000:00:02.0 rcs0: expired last=130:4698, prio=3, hint=3 <idle>-0 0d.s2 1280851192us : __i915_request_unsubmit: 0000:00:02.0 rcs0: fence 66:119966, current 119964 <idle>-0 0d.s2 1280851195us : __i915_request_unsubmit: 0000:00:02.0 rcs0: fence 130:4698, current 4695 <idle>-0 0d.s2 1280851198us : __i915_request_unsubmit: 0000:00:02.0 rcs0: fence 130:4696, current 4695 ^---- Note we unwind 2 requests from the same context <idle>-0 0d.s2 1280851208us : __i915_request_submit: 0000:00:02.0 rcs0: fence 130:4696, current 4695 <idle>-0 0d.s2 1280851213us : __i915_request_submit: 0000:00:02.0 rcs0: fence 134:1508, current 1506 ^---- But to apply the new timeslice, we have to replay the first request before the new client can start -- the unexpected RING_TAIL rewind <idle>-0 0d.s2 1280851219us : trace_ports: 0000:00:02.0 rcs0: submit { 130:4696*, 134:1508 } synmark2-5425 2..s. 1280851239us : process_csb: 0000:00:02.0 rcs0: cs-irq head=5, tail=0 synmark2-5425 2..s. 1280851240us : process_csb: 0000:00:02.0 rcs0: csb[0]: status=0x00008002:0x00000000 ^---- Preemption event for the ELSP update; note the lite-restore synmark2-5425 2..s. 1280851243us : trace_ports: 0000:00:02.0 rcs0: preempted { 130:4698, 66:119966 } synmark2-5425 2..s. 1280851246us : trace_ports: 0000:00:02.0 rcs0: promote { 130:4696*, 134:1508 } synmark2-5425 2.... 1280851462us : __i915_request_commit: 0000:00:02.0 rcs0: fence 130:4700, current 4695 synmark2-5425 2.... 1280852111us : __i915_request_commit: 0000:00:02.0 rcs0: fence 130:4702, current 4695 synmark2-5425 2.Ns1 1280852296us : process_csb: 0000:00:02.0 rcs0: cs-irq head=0, tail=2 synmark2-5425 2.Ns1 1280852297us : process_csb: 0000:00:02.0 rcs0: csb[1]: status=0x00000814:0x00000000 synmark2-5425 2.Ns1 1280852299us : trace_ports: 0000:00:02.0 rcs0: completed { 130:4696!, 134:1508 } synmark2-5425 2.Ns1 1280852301us : process_csb: 0000:00:02.0 rcs0: csb[2]: status=0x00000818:0x00000040 synmark2-5425 2.Ns1 1280852302us : trace_ports: 0000:00:02.0 rcs0: completed { 134:1508, 0:0 } synmark2-5425 2.Ns1 1280852313us : process_csb: process_csb:2336 GEM_BUG_ON(!i915_request_completed(*execlists->active) && !reset_in_progress(execlists)) Fixes: 8ee36e048c98 ("drm/i915/execlists: Minimalistic timeslicing") Referenecs: 82c69bf58650 ("drm/i915/gt: Detect if we miss WaIdleLiteRestore") Signed-off-by: Chris Wilson <chris(a)chris-wilson.co.uk> Cc: Mika Kuoppala <mika.kuoppala(a)linux.intel.com> Reviewed-by: Mika Kuoppala <mika.kuoppala(a)linux.intel.com> Cc: <stable(a)vger.kernel.org> # v5.4+ Link: https://patchwork.freedesktop.org/patch/msgid/20200207211452.2860634-1-chri… (cherry picked from commit 5ba32c7be81e53ea8a27190b0f6be98e6c6779af) --- drivers/gpu/drm/i915/gt/intel_lrc.c | 18 ++++++++---------- drivers/gpu/drm/i915/gt/intel_ring.c | 1 + drivers/gpu/drm/i915/gt/intel_ring.h | 8 ++++++++ drivers/gpu/drm/i915/gt/intel_ring_types.h | 1 + 4 files changed, 18 insertions(+), 10 deletions(-) diff --git a/drivers/gpu/drm/i915/gt/intel_lrc.c b/drivers/gpu/drm/i915/gt/intel_lrc.c index cded952e7ddf1..742c08e79409d 100644 --- a/drivers/gpu/drm/i915/gt/intel_lrc.c +++ b/drivers/gpu/drm/i915/gt/intel_lrc.c @@ -1446,7 +1446,7 @@ static u64 execlists_update_context(struct i915_request *rq) { struct intel_context *ce = rq->context; u64 desc = ce->lrc_desc; - u32 tail; + u32 tail, prev; /* * WaIdleLiteRestore:bdw,skl @@ -1459,9 +1459,15 @@ static u64 execlists_update_context(struct i915_request *rq) * subsequent resubmissions (for lite restore). Should that fail us, * and we try and submit the same tail again, force the context * reload. + * + * If we need to return to a preempted context, we need to skip the + * lite-restore and force it to reload the RING_TAIL. Otherwise, the + * HW has a tendency to ignore us rewinding the TAIL to the end of + * an earlier request. */ tail = intel_ring_set_tail(rq->ring, rq->tail); - if (unlikely(ce->lrc_reg_state[CTX_RING_TAIL] == tail)) + prev = ce->lrc_reg_state[CTX_RING_TAIL]; + if (unlikely(intel_ring_direction(rq->ring, tail, prev) <= 0)) desc |= CTX_DESC_FORCE_RESTORE; ce->lrc_reg_state[CTX_RING_TAIL] = tail; rq->tail = rq->wa_tail; @@ -1979,14 +1985,6 @@ static void execlists_dequeue(struct intel_engine_cs *engine) */ __unwind_incomplete_requests(engine); - /* - * If we need to return to the preempted context, we - * need to skip the lite-restore and force it to - * reload the RING_TAIL. Otherwise, the HW has a - * tendency to ignore us rewinding the TAIL to the - * end of an earlier request. - */ - last->context->lrc_desc |= CTX_DESC_FORCE_RESTORE; last = NULL; } else if (need_timeslice(engine, last) && timer_expired(&engine->execlists.timer)) { diff --git a/drivers/gpu/drm/i915/gt/intel_ring.c b/drivers/gpu/drm/i915/gt/intel_ring.c index 374b28f13ca0b..6ff803f397c4d 100644 --- a/drivers/gpu/drm/i915/gt/intel_ring.c +++ b/drivers/gpu/drm/i915/gt/intel_ring.c @@ -145,6 +145,7 @@ intel_engine_create_ring(struct intel_engine_cs *engine, int size) kref_init(&ring->ref); ring->size = size; + ring->wrap = BITS_PER_TYPE(ring->size) - ilog2(size); /* * Workaround an erratum on the i830 which causes a hang if diff --git a/drivers/gpu/drm/i915/gt/intel_ring.h b/drivers/gpu/drm/i915/gt/intel_ring.h index ea2839d9e0445..5bdce24994aa0 100644 --- a/drivers/gpu/drm/i915/gt/intel_ring.h +++ b/drivers/gpu/drm/i915/gt/intel_ring.h @@ -56,6 +56,14 @@ static inline u32 intel_ring_wrap(const struct intel_ring *ring, u32 pos) return pos & (ring->size - 1); } +static inline int intel_ring_direction(const struct intel_ring *ring, + u32 next, u32 prev) +{ + typecheck(typeof(ring->size), next); + typecheck(typeof(ring->size), prev); + return (next - prev) << ring->wrap; +} + static inline bool intel_ring_offset_valid(const struct intel_ring *ring, unsigned int pos) diff --git a/drivers/gpu/drm/i915/gt/intel_ring_types.h b/drivers/gpu/drm/i915/gt/intel_ring_types.h index d9f17f38e0cce..3cd7fec7fd8d5 100644 --- a/drivers/gpu/drm/i915/gt/intel_ring_types.h +++ b/drivers/gpu/drm/i915/gt/intel_ring_types.h @@ -45,6 +45,7 @@ struct intel_ring { u32 space; u32 size; + u32 wrap; u32 effective_size; }; --------------------------------------------------------------------- Intel Corporation (UK) Limited Registered No. 1134945 (England) Registered Office: Pipers Way, Swindon SN3 1RJ VAT No: 860 2173 47 This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). Any review or distribution by others is strictly prohibited. If you are not the intended recipient, please contact the sender and delete all copies.

5 years, 6 months

1
0
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror