From: Greg Kroah-Hartman Date: Fri, 15 Jan 2021 09:58:38 +0000 (+0100) Subject: 4.14-stable patches X-Git-Tag: v4.4.252~21 X-Git-Url: http://git.ipfire.org/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=eb9f579f3544a592c1450748782f1f46502c1984;p=thirdparty%2Fkernel%2Fstable-queue.git 4.14-stable patches added patches: drm-i915-fix-mismatch-between-misplaced-vma-check-and-vma-insert.patch iio-imu-st_lsm6dsx-fix-edge-trigger-interrupts.patch iio-imu-st_lsm6dsx-flip-irq-return-logic.patch spi-pxa2xx-fix-use-after-free-on-unbind.patch ubifs-wbuf-don-t-leak-kernel-memory-to-flash.patch vmlinux.lds.h-add-pgo-and-autofdo-input-sections.patch x86-resctrl-don-t-move-a-task-to-the-same-resource-group.patch x86-resctrl-use-an-ipi-instead-of-task_work_add-to-update-pqr_assoc-msr.patch --- diff --git a/queue-4.14/drm-i915-fix-mismatch-between-misplaced-vma-check-and-vma-insert.patch b/queue-4.14/drm-i915-fix-mismatch-between-misplaced-vma-check-and-vma-insert.patch new file mode 100644 index 00000000000..735c40182de --- /dev/null +++ b/queue-4.14/drm-i915-fix-mismatch-between-misplaced-vma-check-and-vma-insert.patch @@ -0,0 +1,54 @@ +From foo@baz Fri Jan 15 10:47:51 AM CET 2021 +From: Chris Wilson +Date: Wed, 16 Dec 2020 09:29:51 +0000 +Subject: drm/i915: Fix mismatch between misplaced vma check and vma insert + +From: Chris Wilson + +commit 0e53656ad8abc99e0a80c3de611e593ebbf55829 upstream + +When inserting a VMA, we restrict the placement to the low 4G unless the +caller opts into using the full range. This was done to allow usersapce +the opportunity to transition slowly from a 32b address space, and to +avoid breaking inherent 32b assumptions of some commands. + +However, for insert we limited ourselves to 4G-4K, but on verification +we allowed the full 4G. This causes some attempts to bind a new buffer +to sporadically fail with -ENOSPC, but at other times be bound +successfully. + +commit 48ea1e32c39d ("drm/i915/gen9: Set PIN_ZONE_4G end to 4GB - 1 +page") suggests that there is a genuine problem with stateless addressing +that cannot utilize the last page in 4G and so we purposefully excluded +it. This means that the quick pin pass may cause us to utilize a buggy +placement. + +Reported-by: CQ Tang +Testcase: igt/gem_exec_params/larger-than-life-batch +Fixes: 48ea1e32c39d ("drm/i915/gen9: Set PIN_ZONE_4G end to 4GB - 1 page") +Signed-off-by: Chris Wilson +Cc: CQ Tang +Reviewed-by: CQ Tang +Reviewed-by: Matthew Auld +Cc: # v4.5+ +Link: https://patchwork.freedesktop.org/patch/msgid/20201216092951.7124-1-chris@chris-wilson.co.uk +(cherry picked from commit 5f22cc0b134ab702d7f64b714e26018f7288ffee) +Signed-off-by: Jani Nikula +[sudip: use file from old path] +Signed-off-by: Sudip Mukherjee +Signed-off-by: Greg Kroah-Hartman +--- + drivers/gpu/drm/i915/i915_gem_execbuffer.c | 2 +- + 1 file changed, 1 insertion(+), 1 deletion(-) + +--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c ++++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c +@@ -341,7 +341,7 @@ eb_vma_misplaced(const struct drm_i915_g + return true; + + if (!(flags & EXEC_OBJECT_SUPPORTS_48B_ADDRESS) && +- (vma->node.start + vma->node.size - 1) >> 32) ++ (vma->node.start + vma->node.size + 4095) >> 32) + return true; + + if (flags & __EXEC_OBJECT_NEEDS_MAP && diff --git a/queue-4.14/iio-imu-st_lsm6dsx-fix-edge-trigger-interrupts.patch b/queue-4.14/iio-imu-st_lsm6dsx-fix-edge-trigger-interrupts.patch new file mode 100644 index 00000000000..9037838be5a --- /dev/null +++ b/queue-4.14/iio-imu-st_lsm6dsx-fix-edge-trigger-interrupts.patch @@ -0,0 +1,68 @@ +From foo@baz Fri Jan 15 10:52:27 AM CET 2021 +From: Lorenzo Bianconi +Date: Sat, 14 Nov 2020 19:39:05 +0100 +Subject: iio: imu: st_lsm6dsx: fix edge-trigger interrupts + +From: Lorenzo Bianconi + +commit 3f9bce7a22a3f8ac9d885c9d75bc45569f24ac8b upstream + +If we are using edge IRQs, new samples can arrive while processing +current interrupt since there are no hw guarantees the irq line +stays "low" long enough to properly detect the new interrupt. +In this case the new sample will be missed. +Polling FIFO status register in st_lsm6dsx_handler_thread routine +allow us to read new samples even if the interrupt arrives while +processing previous data and the timeslot where the line is "low" +is too short to be properly detected. + +Fixes: 89ca88a7cdf2 ("iio: imu: st_lsm6dsx: support active-low interrupts") +Fixes: 290a6ce11d93 ("iio: imu: add support to lsm6dsx driver") +Signed-off-by: Lorenzo Bianconi +Link: https://lore.kernel.org/r/5e93cda7dc1e665f5685c53ad8e9ea71dbae782d.1605378871.git.lorenzo@kernel.org +Cc: +Signed-off-by: Jonathan Cameron +[sudip: manual backport to old irq handler path] +Signed-off-by: Sudip Mukherjee +Signed-off-by: Greg Kroah-Hartman +--- + drivers/iio/imu/st_lsm6dsx/st_lsm6dsx_buffer.c | 26 ++++++++++++++++++++----- + 1 file changed, 21 insertions(+), 5 deletions(-) + +--- a/drivers/iio/imu/st_lsm6dsx/st_lsm6dsx_buffer.c ++++ b/drivers/iio/imu/st_lsm6dsx/st_lsm6dsx_buffer.c +@@ -395,13 +395,29 @@ static irqreturn_t st_lsm6dsx_handler_ir + static irqreturn_t st_lsm6dsx_handler_thread(int irq, void *private) + { + struct st_lsm6dsx_hw *hw = private; +- int count; ++ int fifo_len = 0, len; + +- mutex_lock(&hw->fifo_lock); +- count = st_lsm6dsx_read_fifo(hw); +- mutex_unlock(&hw->fifo_lock); ++ /* ++ * If we are using edge IRQs, new samples can arrive while ++ * processing current interrupt since there are no hw ++ * guarantees the irq line stays "low" long enough to properly ++ * detect the new interrupt. In this case the new sample will ++ * be missed. ++ * Polling FIFO status register allow us to read new ++ * samples even if the interrupt arrives while processing ++ * previous data and the timeslot where the line is "low" is ++ * too short to be properly detected. ++ */ ++ do { ++ mutex_lock(&hw->fifo_lock); ++ len = st_lsm6dsx_read_fifo(hw); ++ mutex_unlock(&hw->fifo_lock); + +- return count ? IRQ_HANDLED : IRQ_NONE; ++ if (len > 0) ++ fifo_len += len; ++ } while (len > 0); ++ ++ return fifo_len ? IRQ_HANDLED : IRQ_NONE; + } + + static int st_lsm6dsx_buffer_preenable(struct iio_dev *iio_dev) diff --git a/queue-4.14/iio-imu-st_lsm6dsx-flip-irq-return-logic.patch b/queue-4.14/iio-imu-st_lsm6dsx-flip-irq-return-logic.patch new file mode 100644 index 00000000000..c15136c6ff1 --- /dev/null +++ b/queue-4.14/iio-imu-st_lsm6dsx-flip-irq-return-logic.patch @@ -0,0 +1,31 @@ +From foo@baz Fri Jan 15 10:52:17 AM CET 2021 +From: Sean Nyekjaer +Date: Mon, 15 Jul 2019 09:07:15 +0200 +Subject: iio: imu: st_lsm6dsx: flip irq return logic + +From: Sean Nyekjaer + +commit ec76d918f23034f9f662539ca9c64e2ae3ba9fba upstream + +No need for using reverse logic in the irq return, +fix this by flip things around. + +Signed-off-by: Sean Nyekjaer +Signed-off-by: Jonathan Cameron +Signed-off-by: Sudip Mukherjee +Signed-off-by: Greg Kroah-Hartman +--- + drivers/iio/imu/st_lsm6dsx/st_lsm6dsx_buffer.c | 2 +- + 1 file changed, 1 insertion(+), 1 deletion(-) + +--- a/drivers/iio/imu/st_lsm6dsx/st_lsm6dsx_buffer.c ++++ b/drivers/iio/imu/st_lsm6dsx/st_lsm6dsx_buffer.c +@@ -401,7 +401,7 @@ static irqreturn_t st_lsm6dsx_handler_th + count = st_lsm6dsx_read_fifo(hw); + mutex_unlock(&hw->fifo_lock); + +- return !count ? IRQ_NONE : IRQ_HANDLED; ++ return count ? IRQ_HANDLED : IRQ_NONE; + } + + static int st_lsm6dsx_buffer_preenable(struct iio_dev *iio_dev) diff --git a/queue-4.14/series b/queue-4.14/series index d73947e9600..eaa69e9516c 100644 --- a/queue-4.14/series +++ b/queue-4.14/series @@ -3,3 +3,11 @@ net-cdc_ncm-correct-overhead-in-delayed_ndp_size.patch net-vlan-avoid-leaks-on-register_vlan_dev-failures.patch net-ip-always-refragment-ip-defragmented-packets.patch net-fix-pmtu-check-in-nopmtudisc-mode.patch +x86-resctrl-use-an-ipi-instead-of-task_work_add-to-update-pqr_assoc-msr.patch +x86-resctrl-don-t-move-a-task-to-the-same-resource-group.patch +vmlinux.lds.h-add-pgo-and-autofdo-input-sections.patch +drm-i915-fix-mismatch-between-misplaced-vma-check-and-vma-insert.patch +ubifs-wbuf-don-t-leak-kernel-memory-to-flash.patch +spi-pxa2xx-fix-use-after-free-on-unbind.patch +iio-imu-st_lsm6dsx-flip-irq-return-logic.patch +iio-imu-st_lsm6dsx-fix-edge-trigger-interrupts.patch diff --git a/queue-4.14/spi-pxa2xx-fix-use-after-free-on-unbind.patch b/queue-4.14/spi-pxa2xx-fix-use-after-free-on-unbind.patch new file mode 100644 index 00000000000..c87d6a0bb87 --- /dev/null +++ b/queue-4.14/spi-pxa2xx-fix-use-after-free-on-unbind.patch @@ -0,0 +1,49 @@ +From foo@baz Fri Jan 15 10:50:54 AM CET 2021 +From: Lukas Wunner +Date: Mon, 7 Dec 2020 09:17:05 +0100 +Subject: spi: pxa2xx: Fix use-after-free on unbind + +From: Lukas Wunner + +commit 5626308bb94d9f930aa5f7c77327df4c6daa7759 upstream + +pxa2xx_spi_remove() accesses the driver's private data after calling +spi_unregister_controller() even though that function releases the last +reference on the spi_controller and thereby frees the private data. + +Fix by switching over to the new devm_spi_alloc_master/slave() helper +which keeps the private data accessible until the driver has unbound. + +Fixes: 32e5b57232c0 ("spi: pxa2xx: Fix controller unregister order") +Signed-off-by: Lukas Wunner +Cc: # v2.6.17+: 5e844cc37a5c: spi: Introduce device-managed SPI controller allocation +Cc: # v2.6.17+: 32e5b57232c0: spi: pxa2xx: Fix controller unregister order +Cc: # v2.6.17+ +Link: https://lore.kernel.org/r/5764b04d4a6e43069ebb7808f64c2f774ac6f193.1607286887.git.lukas@wunner.de +Signed-off-by: Mark Brown +[sudip: adjust context] +Signed-off-by: Sudip Mukherjee +Signed-off-by: Greg Kroah-Hartman +--- + drivers/spi/spi-pxa2xx.c | 3 +-- + 1 file changed, 1 insertion(+), 2 deletions(-) + +--- a/drivers/spi/spi-pxa2xx.c ++++ b/drivers/spi/spi-pxa2xx.c +@@ -1660,7 +1660,7 @@ static int pxa2xx_spi_probe(struct platf + return -ENODEV; + } + +- master = spi_alloc_master(dev, sizeof(struct driver_data)); ++ master = devm_spi_alloc_master(dev, sizeof(*drv_data)); + if (!master) { + dev_err(&pdev->dev, "cannot alloc spi_master\n"); + pxa_ssp_free(ssp); +@@ -1841,7 +1841,6 @@ out_error_clock_enabled: + free_irq(ssp->irq, drv_data); + + out_error_master_alloc: +- spi_master_put(master); + pxa_ssp_free(ssp); + return status; + } diff --git a/queue-4.14/ubifs-wbuf-don-t-leak-kernel-memory-to-flash.patch b/queue-4.14/ubifs-wbuf-don-t-leak-kernel-memory-to-flash.patch new file mode 100644 index 00000000000..38959f69f33 --- /dev/null +++ b/queue-4.14/ubifs-wbuf-don-t-leak-kernel-memory-to-flash.patch @@ -0,0 +1,70 @@ +From foo@baz Fri Jan 15 10:49:39 AM CET 2021 +From: Richard Weinberger +Date: Mon, 16 Nov 2020 22:05:30 +0100 +Subject: ubifs: wbuf: Don't leak kernel memory to flash + +From: Richard Weinberger + +commit 20f1431160c6b590cdc269a846fc5a448abf5b98 upstream + +Write buffers use a kmalloc()'ed buffer, they can leak +up to seven bytes of kernel memory to flash if writes are not +aligned. +So use ubifs_pad() to fill these gaps with padding bytes. +This was never a problem while scanning because the scanner logic +manually aligns node lengths and skips over these gaps. + +Cc: +Fixes: 1e51764a3c2ac05a2 ("UBIFS: add new flash file system") +Signed-off-by: Richard Weinberger +Reviewed-by: Zhihao Cheng +Signed-off-by: Richard Weinberger +[sudip: adjust context] +Signed-off-by: Sudip Mukherjee +Signed-off-by: Greg Kroah-Hartman +--- + fs/ubifs/io.c | 13 +++++++++++-- + 1 file changed, 11 insertions(+), 2 deletions(-) + +--- a/fs/ubifs/io.c ++++ b/fs/ubifs/io.c +@@ -331,7 +331,7 @@ void ubifs_pad(const struct ubifs_info * + { + uint32_t crc; + +- ubifs_assert(pad >= 0 && !(pad & 7)); ++ ubifs_assert(pad >= 0); + + if (pad >= UBIFS_PAD_NODE_SZ) { + struct ubifs_ch *ch = buf; +@@ -727,6 +727,10 @@ int ubifs_wbuf_write_nolock(struct ubifs + * write-buffer. + */ + memcpy(wbuf->buf + wbuf->used, buf, len); ++ if (aligned_len > len) { ++ ubifs_assert(aligned_len - len < 8); ++ ubifs_pad(c, wbuf->buf + wbuf->used + len, aligned_len - len); ++ } + + if (aligned_len == wbuf->avail) { + dbg_io("flush jhead %s wbuf to LEB %d:%d", +@@ -819,13 +823,18 @@ int ubifs_wbuf_write_nolock(struct ubifs + } + + spin_lock(&wbuf->lock); +- if (aligned_len) ++ if (aligned_len) { + /* + * And now we have what's left and what does not take whole + * max. write unit, so write it to the write-buffer and we are + * done. + */ + memcpy(wbuf->buf, buf + written, len); ++ if (aligned_len > len) { ++ ubifs_assert(aligned_len - len < 8); ++ ubifs_pad(c, wbuf->buf + len, aligned_len - len); ++ } ++ } + + if (c->leb_size - wbuf->offs >= c->max_write_size) + wbuf->size = c->max_write_size; diff --git a/queue-4.14/vmlinux.lds.h-add-pgo-and-autofdo-input-sections.patch b/queue-4.14/vmlinux.lds.h-add-pgo-and-autofdo-input-sections.patch new file mode 100644 index 00000000000..0ee7a127ce3 --- /dev/null +++ b/queue-4.14/vmlinux.lds.h-add-pgo-and-autofdo-input-sections.patch @@ -0,0 +1,86 @@ +From foo@baz Fri Jan 15 10:46:06 AM CET 2021 +From: Nick Desaulniers +Date: Fri, 21 Aug 2020 12:42:47 -0700 +Subject: vmlinux.lds.h: Add PGO and AutoFDO input sections + +From: Nick Desaulniers + +commit eff8728fe69880d3f7983bec3fb6cea4c306261f upstream. + +Basically, consider .text.{hot|unlikely|unknown}.* part of .text, too. + +When compiling with profiling information (collected via PGO +instrumentations or AutoFDO sampling), Clang will separate code into +.text.hot, .text.unlikely, or .text.unknown sections based on profiling +information. After D79600 (clang-11), these sections will have a +trailing `.` suffix, ie. .text.hot., .text.unlikely., .text.unknown.. + +When using -ffunction-sections together with profiling infomation, +either explicitly (FGKASLR) or implicitly (LTO), code may be placed in +sections following the convention: +.text.hot., .text.unlikely., .text.unknown. +where , , and are functions. (This produces one section +per function; we generally try to merge these all back via linker script +so that we don't have 50k sections). + +For the above cases, we need to teach our linker scripts that such +sections might exist and that we'd explicitly like them grouped +together, otherwise we can wind up with code outside of the +_stext/_etext boundaries that might not be mapped properly for some +architectures, resulting in boot failures. + +If the linker script is not told about possible input sections, then +where the section is placed as output is a heuristic-laiden mess that's +non-portable between linkers (ie. BFD and LLD), and has resulted in many +hard to debug bugs. Kees Cook is working on cleaning this up by adding +--orphan-handling=warn linker flag used in ARCH=powerpc to additional +architectures. In the case of linker scripts, borrowing from the Zen of +Python: explicit is better than implicit. + +Also, ld.bfd's internal linker script considers .text.hot AND +.text.hot.* to be part of .text, as well as .text.unlikely and +.text.unlikely.*. I didn't see support for .text.unknown.*, and didn't +see Clang producing such code in our kernel builds, but I see code in +LLVM that can produce such section names if profiling information is +missing. That may point to a larger issue with generating or collecting +profiles, but I would much rather be safe and explicit than have to +debug yet another issue related to orphan section placement. + +Reported-by: Jian Cai +Suggested-by: Fāng-ruì Sòng +Signed-off-by: Nick Desaulniers +Signed-off-by: Kees Cook +Signed-off-by: Ingo Molnar +Tested-by: Luis Lozano +Tested-by: Manoj Gupta +Acked-by: Kees Cook +Cc: linux-arch@vger.kernel.org +Cc: stable@vger.kernel.org +Link: https://sourceware.org/git/?p=binutils-gdb.git;a=commitdiff;h=add44f8d5c5c05e08b11e033127a744d61c26aee +Link: https://sourceware.org/git/?p=binutils-gdb.git;a=commitdiff;h=1de778ed23ce7492c523d5850c6c6dbb34152655 +Link: https://reviews.llvm.org/D79600 +Link: https://bugs.chromium.org/p/chromium/issues/detail?id=1084760 +Link: https://lore.kernel.org/r/20200821194310.3089815-7-keescook@chromium.org + +Debugged-by: Luis Lozano +[nc: Resolve small conflict due to lack of NOINSTR_TEXT] +Signed-off-by: Nathan Chancellor +Signed-off-by: Greg Kroah-Hartman +--- + include/asm-generic/vmlinux.lds.h | 5 ++++- + 1 file changed, 4 insertions(+), 1 deletion(-) + +--- a/include/asm-generic/vmlinux.lds.h ++++ b/include/asm-generic/vmlinux.lds.h +@@ -459,7 +459,10 @@ + */ + #define TEXT_TEXT \ + ALIGN_FUNCTION(); \ +- *(.text.hot TEXT_MAIN .text.fixup .text.unlikely) \ ++ *(.text.hot .text.hot.*) \ ++ *(TEXT_MAIN .text.fixup) \ ++ *(.text.unlikely .text.unlikely.*) \ ++ *(.text.unknown .text.unknown.*) \ + *(.text..refcount) \ + *(.ref.text) \ + MEM_KEEP(init.text) \ diff --git a/queue-4.14/x86-resctrl-don-t-move-a-task-to-the-same-resource-group.patch b/queue-4.14/x86-resctrl-don-t-move-a-task-to-the-same-resource-group.patch new file mode 100644 index 00000000000..f9a8e9164af --- /dev/null +++ b/queue-4.14/x86-resctrl-don-t-move-a-task-to-the-same-resource-group.patch @@ -0,0 +1,60 @@ +From foo@baz Fri Jan 15 10:40:53 AM CET 2021 +From: Reinette Chatre +Date: Mon, 11 Jan 2021 15:12:58 -0800 +Subject: x86/resctrl: Don't move a task to the same resource group +To: stable@vger.kernel.org, gregkh@linuxfoundation.org +Cc: Fenghua Yu , Shakeel Butt , Reinette Chatre , Borislav Petkov , Tony Luck +Message-ID: <5316a67c4a42a2ddf39df3b9a7c246ab7dff5348.1610394049.git.reinette.chatre@intel.com> + +From: Fenghua Yu + +commit a0195f314a25582b38993bf30db11c300f4f4611 upstream + +Shakeel Butt reported in [1] that a user can request a task to be moved +to a resource group even if the task is already in the group. It just +wastes time to do the move operation which could be costly to send IPI +to a different CPU. + +Add a sanity check to ensure that the move operation only happens when +the task is not already in the resource group. + +[1] https://lore.kernel.org/lkml/CALvZod7E9zzHwenzf7objzGKsdBmVwTgEJ0nPgs0LUFU3SN5Pw@mail.gmail.com/ + +Backporting notes: + +Since upstream commit fa7d949337cc ("x86/resctrl: Rename and move rdt +files to a separate directory"), the file +arch/x86/kernel/cpu/intel_rdt_rdtgroup.c has been renamed and moved to +arch/x86/kernel/cpu/resctrl/rdtgroup.c. +Apply the change against file arch/x86/kernel/cpu/intel_rdt_rdtgroup.c +for older stable trees. + +Fixes: e02737d5b826 ("x86/intel_rdt: Add tasks files") +Reported-by: Shakeel Butt +Signed-off-by: Fenghua Yu +Signed-off-by: Reinette Chatre +Signed-off-by: Borislav Petkov +Reviewed-by: Tony Luck +Cc: stable@vger.kernel.org +Link: https://lkml.kernel.org/r/962ede65d8e95be793cb61102cca37f7bb018e66.1608243147.git.reinette.chatre@intel.com +Signed-off-by: Greg Kroah-Hartman +--- + arch/x86/kernel/cpu/intel_rdt_rdtgroup.c | 7 +++++++ + 1 file changed, 7 insertions(+) + +--- a/arch/x86/kernel/cpu/intel_rdt_rdtgroup.c ++++ b/arch/x86/kernel/cpu/intel_rdt_rdtgroup.c +@@ -432,6 +432,13 @@ static void update_task_closid_rmid(stru + static int __rdtgroup_move_task(struct task_struct *tsk, + struct rdtgroup *rdtgrp) + { ++ /* If the task is already in rdtgrp, no need to move the task. */ ++ if ((rdtgrp->type == RDTCTRL_GROUP && tsk->closid == rdtgrp->closid && ++ tsk->rmid == rdtgrp->mon.rmid) || ++ (rdtgrp->type == RDTMON_GROUP && tsk->rmid == rdtgrp->mon.rmid && ++ tsk->closid == rdtgrp->mon.parent->closid)) ++ return 0; ++ + /* + * Set the task's closid/rmid before the PQR_ASSOC MSR can be + * updated by them. diff --git a/queue-4.14/x86-resctrl-use-an-ipi-instead-of-task_work_add-to-update-pqr_assoc-msr.patch b/queue-4.14/x86-resctrl-use-an-ipi-instead-of-task_work_add-to-update-pqr_assoc-msr.patch new file mode 100644 index 00000000000..07b4642c226 --- /dev/null +++ b/queue-4.14/x86-resctrl-use-an-ipi-instead-of-task_work_add-to-update-pqr_assoc-msr.patch @@ -0,0 +1,216 @@ +From foo@baz Fri Jan 15 10:40:53 AM CET 2021 +From: Reinette Chatre +Date: Mon, 11 Jan 2021 15:12:28 -0800 +Subject: x86/resctrl: Use an IPI instead of task_work_add() to update PQR_ASSOC MSR +To: stable@vger.kernel.org, gregkh@linuxfoundation.org +Cc: Fenghua Yu , Shakeel Butt , Valentin Schneider , Reinette Chatre , Borislav Petkov , Tony Luck , James Morse +Message-ID: <27e975e8975fc10f806d6ede8c28e56872b853cc.1610394049.git.reinette.chatre@intel.com> + +From: Fenghua Yu + +commit ae28d1aae48a1258bd09a6f707ebb4231d79a761 upstream + +Currently, when moving a task to a resource group the PQR_ASSOC MSR is +updated with the new closid and rmid in an added task callback. If the +task is running, the work is run as soon as possible. If the task is not +running, the work is executed later in the kernel exit path when the +kernel returns to the task again. + +Updating the PQR_ASSOC MSR as soon as possible on the CPU a moved task +is running is the right thing to do. Queueing work for a task that is +not running is unnecessary (the PQR_ASSOC MSR is already updated when +the task is scheduled in) and causing system resource waste with the way +in which it is implemented: Work to update the PQR_ASSOC register is +queued every time the user writes a task id to the "tasks" file, even if +the task already belongs to the resource group. + +This could result in multiple pending work items associated with a +single task even if they are all identical and even though only a single +update with most recent values is needed. Specifically, even if a task +is moved between different resource groups while it is sleeping then it +is only the last move that is relevant but yet a work item is queued +during each move. + +This unnecessary queueing of work items could result in significant +system resource waste, especially on tasks sleeping for a long time. +For example, as demonstrated by Shakeel Butt in [1] writing the same +task id to the "tasks" file can quickly consume significant memory. The +same problem (wasted system resources) occurs when moving a task between +different resource groups. + +As pointed out by Valentin Schneider in [2] there is an additional issue +with the way in which the queueing of work is done in that the task_struct +update is currently done after the work is queued, resulting in a race with +the register update possibly done before the data needed by the update is +available. + +To solve these issues, update the PQR_ASSOC MSR in a synchronous way +right after the new closid and rmid are ready during the task movement, +only if the task is running. If a moved task is not running nothing +is done since the PQR_ASSOC MSR will be updated next time the task is +scheduled. This is the same way used to update the register when tasks +are moved as part of resource group removal. + +[1] https://lore.kernel.org/lkml/CALvZod7E9zzHwenzf7objzGKsdBmVwTgEJ0nPgs0LUFU3SN5Pw@mail.gmail.com/ +[2] https://lore.kernel.org/lkml/20201123022433.17905-1-valentin.schneider@arm.com + + [ bp: Massage commit message and drop the two update_task_closid_rmid() + variants. ] + +Backporting notes: + +Since upstream commit fa7d949337cc ("x86/resctrl: Rename and move rdt +files to a separate directory"), the file +arch/x86/kernel/cpu/intel_rdt_rdtgroup.c has been renamed and moved to +arch/x86/kernel/cpu/resctrl/rdtgroup.c. +Apply the change against file arch/x86/kernel/cpu/intel_rdt_rdtgroup.c +for older stable trees. + +Since upstream commit 352940ececaca ("x86/resctrl: Rename the RDT +functions and definitions"), resctrl functions received more generic +names. Specifically related to this backport, intel_rdt_sched_in() +was renamed to rescrl_sched_in(). + +Fixes: e02737d5b826 ("x86/intel_rdt: Add tasks files") +Reported-by: Shakeel Butt +Reported-by: Valentin Schneider +Signed-off-by: Fenghua Yu +Signed-off-by: Reinette Chatre +Signed-off-by: Borislav Petkov +Reviewed-by: Tony Luck +Reviewed-by: James Morse +Reviewed-by: Valentin Schneider +Cc: stable@vger.kernel.org +Link: https://lkml.kernel.org/r/17aa2fb38fc12ce7bb710106b3e7c7b45acb9e94.1608243147.git.reinette.chatre@intel.com +Signed-off-by: Greg Kroah-Hartman +--- + arch/x86/kernel/cpu/intel_rdt_rdtgroup.c | 105 ++++++++++++------------------- + 1 file changed, 42 insertions(+), 63 deletions(-) + +--- a/arch/x86/kernel/cpu/intel_rdt_rdtgroup.c ++++ b/arch/x86/kernel/cpu/intel_rdt_rdtgroup.c +@@ -411,82 +411,61 @@ static void rdtgroup_remove(struct rdtgr + kfree(rdtgrp); + } + +-struct task_move_callback { +- struct callback_head work; +- struct rdtgroup *rdtgrp; +-}; +- +-static void move_myself(struct callback_head *head) ++static void _update_task_closid_rmid(void *task) + { +- struct task_move_callback *callback; +- struct rdtgroup *rdtgrp; +- +- callback = container_of(head, struct task_move_callback, work); +- rdtgrp = callback->rdtgrp; +- + /* +- * If resource group was deleted before this task work callback +- * was invoked, then assign the task to root group and free the +- * resource group. ++ * If the task is still current on this CPU, update PQR_ASSOC MSR. ++ * Otherwise, the MSR is updated when the task is scheduled in. + */ +- if (atomic_dec_and_test(&rdtgrp->waitcount) && +- (rdtgrp->flags & RDT_DELETED)) { +- current->closid = 0; +- current->rmid = 0; +- rdtgroup_remove(rdtgrp); +- } +- +- preempt_disable(); +- /* update PQR_ASSOC MSR to make resource group go into effect */ +- intel_rdt_sched_in(); +- preempt_enable(); ++ if (task == current) ++ intel_rdt_sched_in(); ++} + +- kfree(callback); ++static void update_task_closid_rmid(struct task_struct *t) ++{ ++ if (IS_ENABLED(CONFIG_SMP) && task_curr(t)) ++ smp_call_function_single(task_cpu(t), _update_task_closid_rmid, t, 1); ++ else ++ _update_task_closid_rmid(t); + } + + static int __rdtgroup_move_task(struct task_struct *tsk, + struct rdtgroup *rdtgrp) + { +- struct task_move_callback *callback; +- int ret; +- +- callback = kzalloc(sizeof(*callback), GFP_KERNEL); +- if (!callback) +- return -ENOMEM; +- callback->work.func = move_myself; +- callback->rdtgrp = rdtgrp; +- + /* +- * Take a refcount, so rdtgrp cannot be freed before the +- * callback has been invoked. ++ * Set the task's closid/rmid before the PQR_ASSOC MSR can be ++ * updated by them. ++ * ++ * For ctrl_mon groups, move both closid and rmid. ++ * For monitor groups, can move the tasks only from ++ * their parent CTRL group. + */ +- atomic_inc(&rdtgrp->waitcount); +- ret = task_work_add(tsk, &callback->work, true); +- if (ret) { +- /* +- * Task is exiting. Drop the refcount and free the callback. +- * No need to check the refcount as the group cannot be +- * deleted before the write function unlocks rdtgroup_mutex. +- */ +- atomic_dec(&rdtgrp->waitcount); +- kfree(callback); +- } else { +- /* +- * For ctrl_mon groups move both closid and rmid. +- * For monitor groups, can move the tasks only from +- * their parent CTRL group. +- */ +- if (rdtgrp->type == RDTCTRL_GROUP) { +- tsk->closid = rdtgrp->closid; ++ ++ if (rdtgrp->type == RDTCTRL_GROUP) { ++ tsk->closid = rdtgrp->closid; ++ tsk->rmid = rdtgrp->mon.rmid; ++ } else if (rdtgrp->type == RDTMON_GROUP) { ++ if (rdtgrp->mon.parent->closid == tsk->closid) + tsk->rmid = rdtgrp->mon.rmid; +- } else if (rdtgrp->type == RDTMON_GROUP) { +- if (rdtgrp->mon.parent->closid == tsk->closid) +- tsk->rmid = rdtgrp->mon.rmid; +- else +- ret = -EINVAL; +- } ++ else ++ return -EINVAL; + } +- return ret; ++ ++ /* ++ * Ensure the task's closid and rmid are written before determining if ++ * the task is current that will decide if it will be interrupted. ++ */ ++ barrier(); ++ ++ /* ++ * By now, the task's closid and rmid are set. If the task is current ++ * on a CPU, the PQR_ASSOC MSR needs to be updated to make the resource ++ * group go into effect. If the task is not current, the MSR will be ++ * updated when the task is scheduled in. ++ */ ++ update_task_closid_rmid(tsk); ++ ++ return 0; + } + + static int rdtgroup_task_write_permission(struct task_struct *task,