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

[report] Performance regressions introduced via "cpuidle: menu: Remove iowait influence" on 6.12.y

by ALOK TIWARI

Hi, I’m reporting a performance regression of up to 6% sequential I/O vdbench regression observed on 6.12.y kernel. While running performance benchmarks on v6.12.60 kernel the sequential I/O vdbench metrics are showing a 5-6% performance regression when compared to v6.12.48 Bisect root cause commit ======================== - commit b39b62075ab4 ("cpuidle: menu: Remove iowait influence") Things work fine again when the previously removed performance-multiplier code is added back. Test details ============ The system is connected to a number of disks in disk array using multipathing and directio configuration in the vdbench profile. wd=wd1,sd=sd*,rdpct=0,seekpct=sequential,xfersize=128k rd=128k64T,wd=wd1,iorate=max,elapsed=600,interval=1,warmup=300,threads=64 Thanks, Alok

2 weeks, 5 days

1
0
0 0

[PATCH 6.1] softirq: Add trace points for tasklet entry/exit

by Sumanth Gavini

commit f4bf3ca2e5cba655824b6e0893a98dfb33ed24e5 upstream. Tasklets are supposed to finish their work quickly and should not block the current running process, but it is not guaranteed that they do so. Currently softirq_entry/exit can be used to analyse the total tasklets execution time, but that's not helpful to track individual tasklets execution time. That makes it hard to identify tasklet functions, which take more time than expected. Add tasklet_entry/exit trace point support to track individual tasklet execution. Trivial usage example: # echo 1 > /sys/kernel/debug/tracing/events/irq/tasklet_entry/enable # echo 1 > /sys/kernel/debug/tracing/events/irq/tasklet_exit/enable # cat /sys/kernel/debug/tracing/trace # tracer: nop # # entries-in-buffer/entries-written: 4/4 #P:4 # # _-----=> irqs-off/BH-disabled # / _----=> need-resched # | / _---=> hardirq/softirq # || / _--=> preempt-depth # ||| / _-=> migrate-disable # |||| / delay # TASK-PID CPU# ||||| TIMESTAMP FUNCTION # | | | ||||| | | <idle>-0 [003] ..s1. 314.011428: tasklet_entry: tasklet=0xffffa01ef8db2740 function=tcp_tasklet_func <idle>-0 [003] ..s1. 314.011432: tasklet_exit: tasklet=0xffffa01ef8db2740 function=tcp_tasklet_func <idle>-0 [003] ..s1. 314.017369: tasklet_entry: tasklet=0xffffa01ef8db2740 function=tcp_tasklet_func <idle>-0 [003] ..s1. 314.017371: tasklet_exit: tasklet=0xffffa01ef8db2740 function=tcp_tasklet_func Signed-off-by: Lingutla Chandrasekhar <clingutla(a)codeaurora.org> Signed-off-by: J. Avila <elavila(a)google.com> Signed-off-by: John Stultz <jstultz(a)google.com> Signed-off-by: Thomas Gleixner <tglx(a)linutronix.de> Reviewed-by: Steven Rostedt (Google) <rostedt(a)goodmis.org> Link: https://lore.kernel.org/r/20230407230526.1685443-1-jstultz@google.com [elavila: Port to android-mainline] [jstultz: Rebased to upstream, cut unused trace points, added comments for the tracepoints, reworded commit] Signed-off-by: Sumanth Gavini <sumanth.gavini(a)yahoo.com> --- include/trace/events/irq.h | 47 ++++++++++++++++++++++++++++++++++++++ kernel/softirq.c | 9 ++++++-- 2 files changed, 54 insertions(+), 2 deletions(-) diff --git a/include/trace/events/irq.h b/include/trace/events/irq.h index eeceafaaea4c..a07b4607b663 100644 --- a/include/trace/events/irq.h +++ b/include/trace/events/irq.h @@ -160,6 +160,53 @@ DEFINE_EVENT(softirq, softirq_raise, TP_ARGS(vec_nr) ); +DECLARE_EVENT_CLASS(tasklet, + + TP_PROTO(struct tasklet_struct *t, void *func), + + TP_ARGS(t, func), + + TP_STRUCT__entry( + __field( void *, tasklet) + __field( void *, func) + ), + + TP_fast_assign( + __entry->tasklet = t; + __entry->func = func; + ), + + TP_printk("tasklet=%ps function=%ps", __entry->tasklet, __entry->func) +); + +/** + * tasklet_entry - called immediately before the tasklet is run + * @t: tasklet pointer + * @func: tasklet callback or function being run + * + * Used to find individual tasklet execution time + */ +DEFINE_EVENT(tasklet, tasklet_entry, + + TP_PROTO(struct tasklet_struct *t, void *func), + + TP_ARGS(t, func) +); + +/** + * tasklet_exit - called immediately after the tasklet is run + * @t: tasklet pointer + * @func: tasklet callback or function being run + * + * Used to find individual tasklet execution time + */ +DEFINE_EVENT(tasklet, tasklet_exit, + + TP_PROTO(struct tasklet_struct *t, void *func), + + TP_ARGS(t, func) +); + #endif /* _TRACE_IRQ_H */ /* This part must be outside protection */ diff --git a/kernel/softirq.c b/kernel/softirq.c index 9ab5ca399a99..fadc6bbda27b 100644 --- a/kernel/softirq.c +++ b/kernel/softirq.c @@ -822,10 +822,15 @@ static void tasklet_action_common(struct softirq_action *a, if (tasklet_trylock(t)) { if (!atomic_read(&t->count)) { if (tasklet_clear_sched(t)) { - if (t->use_callback) + if (t->use_callback) { + trace_tasklet_entry(t, t->callback); t->callback(t); - else + trace_tasklet_exit(t, t->callback); + } else { + trace_tasklet_entry(t, t->func); t->func(t->data); + trace_tasklet_exit(t, t->func); + } } tasklet_unlock(t); continue; -- 2.43.0

2 weeks, 5 days

6
11
0 0

[PATCH v2] gpio: regmap: Fix memleak in gpio_remap_register

by Wentao Guan

We should call gpiochip_remove(chip) to free the resource alloced by gpiochip_add_data(chip, gpio) after the err path. Fixes: 553b75d4bfe9 ("gpio: regmap: Allow to allocate regmap-irq device") CC: stable(a)vger.kernel.org Co-developed-by: WangYuli <wangyl5933(a)chinaunicom.cn> Signed-off-by: WangYuli <wangyl5933(a)chinaunicom.cn> Signed-off-by: Wentao Guan <guanwentao(a)uniontech.com> --- changelog in v2: 1. format commit message. 2. rebase to mainline now. --- --- drivers/gpio/gpio-regmap.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpio/gpio-regmap.c b/drivers/gpio/gpio-regmap.c index f4267af00027e..c64805dcb9f88 100644 --- a/drivers/gpio/gpio-regmap.c +++ b/drivers/gpio/gpio-regmap.c @@ -328,7 +328,7 @@ struct gpio_regmap *gpio_regmap_register(const struct gpio_regmap_config *config config->regmap_irq_line, config->regmap_irq_flags, 0, config->regmap_irq_chip, &gpio->irq_chip_data); if (ret) - goto err_free_bitmap; + goto err_remove_gpiochip; irq_domain = regmap_irq_get_domain(gpio->irq_chip_data); } else base-commit: 3f9f0252130e7dd60d41be0802bf58f6471c691d -- 2.20.1

2 weeks, 5 days

2
3
0 0

FAILED: patch "[PATCH] drm, fbcon, vga_switcheroo: Avoid race condition in fbcon" failed to apply to 6.12-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 6.12-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-6.12.y git checkout FETCH_HEAD git cherry-pick -x eb76d0f5553575599561010f24c277cc5b31d003 # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2025120119-edgy-recycled-bcfe@gregkh' --subject-prefix 'PATCH 6.12.y' HEAD^.. Possible dependencies: thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From eb76d0f5553575599561010f24c277cc5b31d003 Mon Sep 17 00:00:00 2001 From: Thomas Zimmermann <tzimmermann(a)suse.de> Date: Wed, 5 Nov 2025 17:14:49 +0100 Subject: [PATCH] drm, fbcon, vga_switcheroo: Avoid race condition in fbcon setup Protect vga_switcheroo_client_fb_set() with console lock. Avoids OOB access in fbcon_remap_all(). Without holding the console lock the call races with switching outputs. VGA switcheroo calls fbcon_remap_all() when switching clients. The fbcon function uses struct fb_info.node, which is set by register_framebuffer(). As the fb-helper code currently sets up VGA switcheroo before registering the framebuffer, the value of node is -1 and therefore not a legal value. For example, fbcon uses the value within set_con2fb_map() [1] as an index into an array. Moving vga_switcheroo_client_fb_set() after register_framebuffer() can result in VGA switching that does not switch fbcon correctly. Therefore move vga_switcheroo_client_fb_set() under fbcon_fb_registered(), which already holds the console lock. Fbdev calls fbcon_fb_registered() from within register_framebuffer(). Serializes the helper with VGA switcheroo's call to fbcon_remap_all(). Although vga_switcheroo_client_fb_set() takes an instance of struct fb_info as parameter, it really only needs the contained fbcon state. Moving the call to fbcon initialization is therefore cleaner than before. Only amdgpu, i915, nouveau and radeon support vga_switcheroo. For all other drivers, this change does nothing. Signed-off-by: Thomas Zimmermann <tzimmermann(a)suse.de> Link: https://elixir.bootlin.com/linux/v6.17/source/drivers/video/fbdev/core/fbco… # [1] Fixes: 6a9ee8af344e ("vga_switcheroo: initial implementation (v15)") Acked-by: Javier Martinez Canillas <javierm(a)redhat.com> Acked-by: Alex Deucher <alexander.deucher(a)amd.com> Cc: dri-devel(a)lists.freedesktop.org Cc: nouveau(a)lists.freedesktop.org Cc: amd-gfx(a)lists.freedesktop.org Cc: linux-fbdev(a)vger.kernel.org Cc: <stable(a)vger.kernel.org> # v2.6.34+ Link: https://patch.msgid.link/20251105161549.98836-1-tzimmermann@suse.de diff --git a/drivers/gpu/drm/drm_fb_helper.c b/drivers/gpu/drm/drm_fb_helper.c index 11a5b60cb9ce..0b3ee008523d 100644 --- a/drivers/gpu/drm/drm_fb_helper.c +++ b/drivers/gpu/drm/drm_fb_helper.c @@ -31,9 +31,7 @@ #include <linux/console.h> #include <linux/export.h> -#include <linux/pci.h> #include <linux/sysrq.h> -#include <linux/vga_switcheroo.h> #include <drm/drm_atomic.h> #include <drm/drm_drv.h> @@ -566,11 +564,6 @@ EXPORT_SYMBOL(drm_fb_helper_release_info); */ void drm_fb_helper_unregister_info(struct drm_fb_helper *fb_helper) { - struct fb_info *info = fb_helper->info; - struct device *dev = info->device; - - if (dev_is_pci(dev)) - vga_switcheroo_client_fb_set(to_pci_dev(dev), NULL); unregister_framebuffer(fb_helper->info); } EXPORT_SYMBOL(drm_fb_helper_unregister_info); @@ -1632,7 +1625,6 @@ static int drm_fb_helper_single_fb_probe(struct drm_fb_helper *fb_helper) struct drm_client_dev *client = &fb_helper->client; struct drm_device *dev = fb_helper->dev; struct drm_fb_helper_surface_size sizes; - struct fb_info *info; int ret; if (drm_WARN_ON(dev, !dev->driver->fbdev_probe)) @@ -1653,12 +1645,6 @@ static int drm_fb_helper_single_fb_probe(struct drm_fb_helper *fb_helper) strcpy(fb_helper->fb->comm, "[fbcon]"); - info = fb_helper->info; - - /* Set the fb info for vgaswitcheroo clients. Does nothing otherwise. */ - if (dev_is_pci(info->device)) - vga_switcheroo_client_fb_set(to_pci_dev(info->device), info); - return 0; } diff --git a/drivers/video/fbdev/core/fbcon.c b/drivers/video/fbdev/core/fbcon.c index 9bd3c3814b5c..e7e07eb2142e 100644 --- a/drivers/video/fbdev/core/fbcon.c +++ b/drivers/video/fbdev/core/fbcon.c @@ -66,6 +66,7 @@ #include <linux/string.h> #include <linux/kd.h> #include <linux/panic.h> +#include <linux/pci.h> #include <linux/printk.h> #include <linux/slab.h> #include <linux/fb.h> @@ -78,6 +79,7 @@ #include <linux/interrupt.h> #include <linux/crc32.h> /* For counting font checksums */ #include <linux/uaccess.h> +#include <linux/vga_switcheroo.h> #include <asm/irq.h> #include "fbcon.h" @@ -2899,6 +2901,9 @@ void fbcon_fb_unregistered(struct fb_info *info) console_lock(); + if (info->device && dev_is_pci(info->device)) + vga_switcheroo_client_fb_set(to_pci_dev(info->device), NULL); + fbcon_registered_fb[info->node] = NULL; fbcon_num_registered_fb--; @@ -3032,6 +3037,10 @@ static int do_fb_registered(struct fb_info *info) } } + /* Set the fb info for vga_switcheroo clients. Does nothing otherwise. */ + if (info->device && dev_is_pci(info->device)) + vga_switcheroo_client_fb_set(to_pci_dev(info->device), info); + return ret; }

2 weeks, 5 days

4
5
0 0

LDI Show 2025 Data

by Rachel Porter

Hi, Ahead of LDI Show 2025 (Dec 3–9, Las Vegas), we have access to a verified audience of 16,594 and a strong exhibitor roster of 312 top suppliers and vendors. This includes lighting & audio equipment manufacturers, staging/rigging vendors, media-server / projection / video providers, production houses, rental companies, sound & lighting engineers, stage managers ,other live-event technology stakeholders and many. If interested, reply “Send Pricing” to receive the details. Best regards, Rachel Porter Head of Client Relations To opt-out, reply “Not Interested”.

2 weeks, 5 days

1
0
0 0

[PATCH] perf build: build BPF skeletons with fPIC

by Jon Kohler

Fix Makefile.perf to ensure that bpf skeletons are built with fPIC. When building with BUILD_BPF_SKEL=1, bpf_skel's was not getting built with fPIC, seeing compilation failures like: /usr/bin/ld: /builddir/.../tools/perf/util/bpf_skel/.tmp/bootstrap/main.o: relocation R_X86_64_32 against `.rodata.str1.8' can not be used when making a PIE object; recompile with -fPIE Bisected down to 6.18 commit a39516805992 ("tools build: Don't assume libtracefs-devel is always available"). Fixes: a39516805992 ("tools build: Don't assume libtracefs-devel is always available") Cc: stable(a)vger.kernel.org Signed-off-by: Jon Kohler <jon(a)nutanix.com> --- tools/perf/Makefile.perf | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/perf/Makefile.perf b/tools/perf/Makefile.perf index 02f87c49801f..4557c2e89e88 100644 --- a/tools/perf/Makefile.perf +++ b/tools/perf/Makefile.perf @@ -1211,7 +1211,7 @@ endif $(BPFTOOL): | $(SKEL_TMP_OUT) $(Q)CFLAGS= $(MAKE) -C ../bpf/bpftool \ - OUTPUT=$(SKEL_TMP_OUT)/ bootstrap + EXTRA_CFLAGS="-fPIC" OUTPUT=$(SKEL_TMP_OUT)/ bootstrap # Paths to search for a kernel to generate vmlinux.h from. VMLINUX_BTF_ELF_PATHS ?= $(if $(O),$(O)/vmlinux) \ -- 2.43.0

2 weeks, 5 days

2
2
0 0

[PATCH v2 1/2] kasan: Refactor pcpu kasan vmalloc unpoison

by Maciej Wieczor-Retman

From: Maciej Wieczor-Retman <maciej.wieczor-retman(a)intel.com> A KASAN tag mismatch, possibly causing a kernel panic, can be observed on systems with a tag-based KASAN enabled and with multiple NUMA nodes. It was reported on arm64 and reproduced on x86. It can be explained in the following points: 1. There can be more than one virtual memory chunk. 2. Chunk's base address has a tag. 3. The base address points at the first chunk and thus inherits the tag of the first chunk. 4. The subsequent chunks will be accessed with the tag from the first chunk. 5. Thus, the subsequent chunks need to have their tag set to match that of the first chunk. Refactor code by reusing __kasan_unpoison_vmalloc in a new helper in preparation for the actual fix. Changelog v1 (after splitting of from the KASAN series): - Rewrite first paragraph of the patch message to point at the user impact of the issue. - Move helper to common.c so it can be compiled in all KASAN modes. Fixes: 1d96320f8d53 ("kasan, vmalloc: add vmalloc tagging for SW_TAGS") Cc: <stable(a)vger.kernel.org> # 6.1+ Signed-off-by: Maciej Wieczor-Retman <maciej.wieczor-retman(a)intel.com> --- Changelog v2: - Redo the whole patch so it's an actual refactor. include/linux/kasan.h | 16 +++++++++++++--- mm/kasan/common.c | 17 +++++++++++++++++ mm/kasan/hw_tags.c | 15 +++++++++++++-- mm/kasan/shadow.c | 16 ++++++++++++++-- mm/vmalloc.c | 4 +--- 5 files changed, 58 insertions(+), 10 deletions(-) diff --git a/include/linux/kasan.h b/include/linux/kasan.h index d12e1a5f5a9a..4a3d3dba9764 100644 --- a/include/linux/kasan.h +++ b/include/linux/kasan.h @@ -595,14 +595,14 @@ static inline void kasan_release_vmalloc(unsigned long start, #endif /* CONFIG_KASAN_GENERIC || CONFIG_KASAN_SW_TAGS */ -void *__kasan_unpoison_vmalloc(const void *start, unsigned long size, - kasan_vmalloc_flags_t flags); +void *__kasan_random_unpoison_vmalloc(const void *start, unsigned long size, + kasan_vmalloc_flags_t flags); static __always_inline void *kasan_unpoison_vmalloc(const void *start, unsigned long size, kasan_vmalloc_flags_t flags) { if (kasan_enabled()) - return __kasan_unpoison_vmalloc(start, size, flags); + return __kasan_random_unpoison_vmalloc(start, size, flags); return (void *)start; } @@ -614,6 +614,11 @@ static __always_inline void kasan_poison_vmalloc(const void *start, __kasan_poison_vmalloc(start, size); } +void *__kasan_unpoison_vmap_areas(void *addr, unsigned long size, + kasan_vmalloc_flags_t flags, u8 tag); +void kasan_unpoison_vmap_areas(struct vm_struct **vms, int nr_vms, + kasan_vmalloc_flags_t flags); + #else /* CONFIG_KASAN_VMALLOC */ static inline void kasan_populate_early_vm_area_shadow(void *start, @@ -638,6 +643,11 @@ static inline void *kasan_unpoison_vmalloc(const void *start, static inline void kasan_poison_vmalloc(const void *start, unsigned long size) { } +static __always_inline void +kasan_unpoison_vmap_areas(struct vm_struct **vms, int nr_vms, + kasan_vmalloc_flags_t flags) +{ } + #endif /* CONFIG_KASAN_VMALLOC */ #if (defined(CONFIG_KASAN_GENERIC) || defined(CONFIG_KASAN_SW_TAGS)) && \ diff --git a/mm/kasan/common.c b/mm/kasan/common.c index d4c14359feaf..7884ea7d13f9 100644 --- a/mm/kasan/common.c +++ b/mm/kasan/common.c @@ -28,6 +28,7 @@ #include <linux/string.h> #include <linux/types.h> #include <linux/bug.h> +#include <linux/vmalloc.h> #include "kasan.h" #include "../slab.h" @@ -582,3 +583,19 @@ bool __kasan_check_byte(const void *address, unsigned long ip) } return true; } + +#ifdef CONFIG_KASAN_VMALLOC +void kasan_unpoison_vmap_areas(struct vm_struct **vms, int nr_vms, + kasan_vmalloc_flags_t flags) +{ + unsigned long size; + void *addr; + int area; + + for (area = 0 ; area < nr_vms ; area++) { + size = vms[area]->size; + addr = vms[area]->addr; + vms[area]->addr = __kasan_unpoison_vmap_areas(addr, size, flags); + } +} +#endif diff --git a/mm/kasan/hw_tags.c b/mm/kasan/hw_tags.c index 1c373cc4b3fa..4b7936a2bd6f 100644 --- a/mm/kasan/hw_tags.c +++ b/mm/kasan/hw_tags.c @@ -316,8 +316,8 @@ static void init_vmalloc_pages(const void *start, unsigned long size) } } -void *__kasan_unpoison_vmalloc(const void *start, unsigned long size, - kasan_vmalloc_flags_t flags) +static void *__kasan_unpoison_vmalloc(const void *start, unsigned long size, + kasan_vmalloc_flags_t flags) { u8 tag; unsigned long redzone_start, redzone_size; @@ -387,6 +387,12 @@ void *__kasan_unpoison_vmalloc(const void *start, unsigned long size, return (void *)start; } +void *__kasan_random_unpoison_vmalloc(const void *start, unsigned long size, + kasan_vmalloc_flags_t flags) +{ + return __kasan_unpoison_vmalloc(start, size, flags); +} + void __kasan_poison_vmalloc(const void *start, unsigned long size) { /* @@ -396,6 +402,11 @@ void __kasan_poison_vmalloc(const void *start, unsigned long size) */ } +void *__kasan_unpoison_vmap_areas(void *addr, unsigned long size, + kasan_vmalloc_flags_t flags, u8 tag) +{ + return __kasan_unpoison_vmalloc(addr, size, flags); +} #endif void kasan_enable_hw_tags(void) diff --git a/mm/kasan/shadow.c b/mm/kasan/shadow.c index 5d2a876035d6..0a8d8bf6e9cf 100644 --- a/mm/kasan/shadow.c +++ b/mm/kasan/shadow.c @@ -624,8 +624,8 @@ void kasan_release_vmalloc(unsigned long start, unsigned long end, } } -void *__kasan_unpoison_vmalloc(const void *start, unsigned long size, - kasan_vmalloc_flags_t flags) +static void *__kasan_unpoison_vmalloc(const void *start, unsigned long size, + kasan_vmalloc_flags_t flags) { /* * Software KASAN modes unpoison both VM_ALLOC and non-VM_ALLOC @@ -653,6 +653,18 @@ void *__kasan_unpoison_vmalloc(const void *start, unsigned long size, return (void *)start; } +void *__kasan_random_unpoison_vmalloc(const void *start, unsigned long size, + kasan_vmalloc_flags_t flags) +{ + return __kasan_unpoison_vmalloc(start, size, flags); +} + +void *__kasan_unpoison_vmap_areas(void *addr, unsigned long size, + kasan_vmalloc_flags_t flags, u8 tag) +{ + return __kasan_unpoison_vmalloc(addr, size, flags); +} + /* * Poison the shadow for a vmalloc region. Called as part of the * freeing process at the time the region is freed. diff --git a/mm/vmalloc.c b/mm/vmalloc.c index 798b2ed21e46..32ecdb8cd4b8 100644 --- a/mm/vmalloc.c +++ b/mm/vmalloc.c @@ -4870,9 +4870,7 @@ struct vm_struct **pcpu_get_vm_areas(const unsigned long *offsets, * With hardware tag-based KASAN, marking is skipped for * non-VM_ALLOC mappings, see __kasan_unpoison_vmalloc(). */ - for (area = 0; area < nr_vms; area++) - vms[area]->addr = kasan_unpoison_vmalloc(vms[area]->addr, - vms[area]->size, KASAN_VMALLOC_PROT_NORMAL); + kasan_unpoison_vmap_areas(vms, nr_vms, KASAN_VMALLOC_PROT_NORMAL); kfree(vas); return vms; -- 2.52.0

2 weeks, 5 days

3
2
0 0

[PATCH] comedi: Fix getting range information for subdevices 16 to 255

by Ian Abbott

The `COMEDI_RANGEINFO` ioctl does not work properly for subdevice indices above 15. Currently, the only in-tree COMEDI drivers that support more than 16 subdevices are the "8255" driver and the "comedi_bond" driver. Making the ioctl work for subdevice indices up to 255 is achievable. It needs minor changes to the handling of the `COMEDI_RANGEINFO` and `COMEDI_CHANINFO` ioctls that should be mostly harmless to user-space, apart from making them less broken. Details follow... The `COMEDI_RANGEINFO` ioctl command gets the list of supported ranges (usually with units of volts or milliamps) for a COMEDI subdevice or channel. (Only some subdevices have per-channel range tables, indicated by the `SDF_RANGETYPE` flag in the subdevice information.) It uses a `range_type` value and a user-space pointer, both supplied by user-space, but the `range_type` value should match what was obtained using the `COMEDI_CHANINFO` ioctl (if the subdevice has per-channel range tables) or `COMEDI_SUBDINFO` ioctl (if the subdevice uses a single range table for all channels). Bits 15 to 0 of the `range_type` value contain the length of the range table, which is the only part that user-space should care about (so it can use a suitably sized buffer to fetch the range table). Bits 23 to 16 store the channel index, which is assumed to be no more than 255 if the subdevice has per-channel range tables, and is set to 0 if the subdevice has a single range table. For `range_type` values produced by the `COMEDI_SUBDINFO` ioctl, bits 31 to 24 contain the subdevice index, which is assumed to be no more than 255. But for `range_type` values produced by the `COMEDI_CHANINFO` ioctl, bits 27 to 24 contain the subdevice index, which is assumed to be no more than 15, and bits 31 to 28 contain the COMEDI device's minor device number for some unknown reason lost in the mists of time. The `COMEDI_RANGEINFO` ioctl extract the length from bits 15 to 0 of the user-supplied `range_type` value, extracts the channel index from bits 23 to 16 (only used if the subdevice has per-channel range tables), extracts the subdevice index from bits 27 to 24, and ignores bits 31 to 28. So for subdevice indices 16 to 255, the `COMEDI_SUBDINFO` or `COMEDI_CHANINFO` ioctl will report a `range_type` value that doesn't work with the `COMEDI_RANGEINFO` ioctl. It will either get the range table for the subdevice index modulo 16, or will fail with `-EINVAL`. To fix this, always use bits 31 to 24 of the `range_type` value to hold the subdevice index (assumed to be no more than 255). This affects the `COMEDI_CHANINFO` and `COMEDI_RANGEINFO` ioctls. There should not be anything in user-space that depends on the old, broken usage, although it may now see different values in bits 31 to 28 of the `range_type` values reported by the `COMEDI_CHANINFO` ioctl for subdevices that have per-channel subdevices. User-space should not be trying to decode bits 31 to 16 of the `range_type` values anyway. Fixes: ed9eccbe8970 ("Staging: add comedi core") Cc: <stable(a)vger.kernel.org> #5.17+ Signed-off-by: Ian Abbott <abbotti(a)mev.co.uk> --- Backports needed for kernels prior to 5.17 due to .c and .h files moving to different parts of the tree. --- drivers/comedi/comedi_fops.c | 2 +- drivers/comedi/range.c | 2 +- include/uapi/linux/comedi.h | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/comedi/comedi_fops.c b/drivers/comedi/comedi_fops.c index 657c98cd723e..2c3eb9e89571 100644 --- a/drivers/comedi/comedi_fops.c +++ b/drivers/comedi/comedi_fops.c @@ -1155,7 +1155,7 @@ static int do_chaninfo_ioctl(struct comedi_device *dev, for (i = 0; i < s->n_chan; i++) { int x; - x = (dev->minor << 28) | (it->subdev << 24) | (i << 16) | + x = (it->subdev << 24) | (i << 16) | (s->range_table_list[i]->length); if (put_user(x, it->rangelist + i)) return -EFAULT; diff --git a/drivers/comedi/range.c b/drivers/comedi/range.c index 8f43cf88d784..5b8f662365e3 100644 --- a/drivers/comedi/range.c +++ b/drivers/comedi/range.c @@ -52,7 +52,7 @@ int do_rangeinfo_ioctl(struct comedi_device *dev, const struct comedi_lrange *lr; struct comedi_subdevice *s; - subd = (it->range_type >> 24) & 0xf; + subd = (it->range_type >> 24) & 0xff; chan = (it->range_type >> 16) & 0xff; if (!dev->attached) diff --git a/include/uapi/linux/comedi.h b/include/uapi/linux/comedi.h index 7314e5ee0a1e..798ec9a39e12 100644 --- a/include/uapi/linux/comedi.h +++ b/include/uapi/linux/comedi.h @@ -640,7 +640,7 @@ struct comedi_chaninfo { /** * struct comedi_rangeinfo - used to retrieve the range table for a channel - * @range_type: Encodes subdevice index (bits 27:24), channel index + * @range_type: Encodes subdevice index (bits 31:24), channel index * (bits 23:16) and range table length (bits 15:0). * @range_ptr: Pointer to array of @struct comedi_krange to be filled * in with the range table for the channel or subdevice. -- 2.51.0

2 weeks, 5 days

1
0
0 0

Live Design 2025:Contacts

by Melissa Underwood

Hi, As you have been an exhibitor at Live Design International 2025 We have the final updated attendee list of verified contacts 15,805 of people including last-minute registers and walk-ins to the show, and all are confirmed Opt-in and Verified contacts Post Thanksgiving Day Special : 20% reduction on our compliant contact data. If you'd like more details, just let me know or reply "Send me pricing." Best regards, Melissa Underwood Sr. Marketing Manager If you prefer not to receive updates, reply "Not Interested."

2 weeks, 5 days

1
0
0 0

[PATCH v3] KVM: x86: Add x2APIC "features" to control EOI broadcast suppression

by Khushit Shah

Add two flags for KVM_CAP_X2APIC_API to allow userspace to control support for Suppress EOI Broadcasts, which KVM completely mishandles. When x2APIC support was first added, KVM incorrectly advertised and "enabled" Suppress EOI Broadcast, without fully supporting the I/O APIC side of the equation, i.e. without adding directed EOI to KVM's in-kernel I/O APIC. That flaw was carried over to split IRQCHIP support, i.e. KVM advertised support for Suppress EOI Broadcasts irrespective of whether or not the userspace I/O APIC implementation supported directed EOIs. Even worse, KVM didn't actually suppress EOI broadcasts, i.e. userspace VMMs without support for directed EOI came to rely on the "spurious" broadcasts. KVM "fixed" the in-kernel I/O APIC implementation by completely disabling support for Suppress EOI Broadcasts in commit 0bcc3fb95b97 ("KVM: lapic: stop advertising DIRECTED_EOI when in-kernel IOAPIC is in use"), but didn't do anything to remedy userspace I/O APIC implementations. KVM's bogus handling of Suppress EOI Broadcast is problematic when the guest relies on interrupts being masked in the I/O APIC until well after the initial local APIC EOI. E.g. Windows with Credential Guard enabled handles interrupts in the following order: 1. Interrupt for L2 arrives. 2. L1 APIC EOIs the interrupt. 3. L1 resumes L2 and injects the interrupt. 4. L2 EOIs after servicing. 5. L1 performs the I/O APIC EOI. Because KVM EOIs the I/O APIC at step #2, the guest can get an interrupt storm, e.g. if the IRQ line is still asserted and userspace reacts to the EOI by re-injecting the IRQ, because the guest doesn't de-assert the line until step #4, and doesn't expect the interrupt to be re-enabled until step #5. Unfortunately, simply "fixing" the bug isn't an option, as KVM has no way of knowing if the userspace I/O APIC supports directed EOIs, i.e. suppressing EOI broadcasts would result in interrupts being stuck masked in the userspace I/O APIC due to step #5 being ignored by userspace. And fully disabling support for Suppress EOI Broadcast is also undesirable, as picking up the fix would require a guest reboot, *and* more importantly would change the virtual CPU model exposed to the guest without any buy-in from userspace. Add two flags to allow userspace to choose exactly how to solve the immediate issue, and in the long term to allow userspace to control the virtual CPU model that is exposed to the guest (KVM should never have enabled support for Suppress EOI Broadcast without a userspace opt-in). Note, Suppress EOI Broadcasts is defined only in Intel's SDM, not in AMD's APM. But the bit is writable on some AMD CPUs, e.g. Turin, and KVM's ABI is to support Directed EOI (KVM's name) irrespective of guest CPU vendor. Fixes: 7543a635aa09 ("KVM: x86: Add KVM exit for IOAPIC EOIs") Closes: https://lore.kernel.org/kvm/7D497EF1-607D-4D37-98E7-DAF95F099342@nutanix.com Cc: stable(a)vger.kernel.org Co-developed-by: Sean Christopherson <seanjc(a)google.com> Signed-off-by: Sean Christopherson <seanjc(a)google.com> Signed-off-by: Khushit Shah <khushit.shah(a)nutanix.com> --- v3: - Resend with correct patch contents; v2 mail formatting was broken. No functional changes beyond the naming and grammar feedback from v1. Testing: I ran the tests with QEMU 9.1 and a 6.12 kernel with the patch applied. - With an unmodified QEMU build, KVM's LAPIC SEOIB behavior remains unchanged. - Invoking the x2APIC API with KVM_X2APIC_API_DISABLE_IGNORE_SUPPRESS_EOI_BROADCAST_QUIRK correctly suppresses LAPIC -> IOAPIC EOI broadcasts (verified via KVM tracepoints). - Invoking the x2APIC API with KVM_X2APIC_API_DISABLE_SUPPRESS_EOI_BROADCAST results in SEOIB not being advertised to the guest, as expected (confirmed by checking the LAPIC LVR value inside the guest). I'll send the corresponding QEMU-side patch shortly. --- Documentation/virt/kvm/api.rst | 14 ++++++++++++-- arch/x86/include/asm/kvm_host.h | 2 ++ arch/x86/include/uapi/asm/kvm.h | 6 ++++-- arch/x86/kvm/lapic.c | 13 +++++++++++++ arch/x86/kvm/x86.c | 12 +++++++++--- 5 files changed, 40 insertions(+), 7 deletions(-) diff --git a/Documentation/virt/kvm/api.rst b/Documentation/virt/kvm/api.rst index 57061fa29e6a..4141d2bd8156 100644 --- a/Documentation/virt/kvm/api.rst +++ b/Documentation/virt/kvm/api.rst @@ -7800,8 +7800,10 @@ Will return -EBUSY if a VCPU has already been created. Valid feature flags in args[0] are:: - #define KVM_X2APIC_API_USE_32BIT_IDS (1ULL << 0) - #define KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK (1ULL << 1) + #define KVM_X2APIC_API_USE_32BIT_IDS (1ULL << 0) + #define KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK (1ULL << 1) + #define KVM_X2APIC_API_DISABLE_IGNORE_SUPPRESS_EOI_BROADCAST_QUIRK (1ULL << 2) + #define KVM_X2APIC_API_DISABLE_SUPPRESS_EOI_BROADCAST (1ULL << 3) Enabling KVM_X2APIC_API_USE_32BIT_IDS changes the behavior of KVM_SET_GSI_ROUTING, KVM_SIGNAL_MSI, KVM_SET_LAPIC, and KVM_GET_LAPIC, @@ -7814,6 +7816,14 @@ as a broadcast even in x2APIC mode in order to support physical x2APIC without interrupt remapping. This is undesirable in logical mode, where 0xff represents CPUs 0-7 in cluster 0. +Setting KVM_X2APIC_API_DISABLE_IGNORE_SUPPRESS_EOI_BROADCAST_QUIRK overrides +KVM's quirky behavior of not actually suppressing EOI broadcasts for split IRQ +chips when support for Suppress EOI Broadcasts is advertised to the guest. + +Setting KVM_X2APIC_API_DISABLE_SUPPRESS_EOI_BROADCAST disables support for +Suppress EOI Broadcasts entirely, i.e. instructs KVM to NOT advertise support +to the guest and thus disallow enabling EOI broadcast suppression in SPIV. + 7.8 KVM_CAP_S390_USER_INSTR0 ---------------------------- diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h index 48598d017d6f..f6fdc0842c05 100644 --- a/arch/x86/include/asm/kvm_host.h +++ b/arch/x86/include/asm/kvm_host.h @@ -1480,6 +1480,8 @@ struct kvm_arch { bool x2apic_format; bool x2apic_broadcast_quirk_disabled; + bool disable_ignore_suppress_eoi_broadcast_quirk; + bool x2apic_disable_suppress_eoi_broadcast; bool has_mapped_host_mmio; bool guest_can_read_msr_platform_info; diff --git a/arch/x86/include/uapi/asm/kvm.h b/arch/x86/include/uapi/asm/kvm.h index d420c9c066d4..7255385b6d80 100644 --- a/arch/x86/include/uapi/asm/kvm.h +++ b/arch/x86/include/uapi/asm/kvm.h @@ -913,8 +913,10 @@ struct kvm_sev_snp_launch_finish { __u64 pad1[4]; }; -#define KVM_X2APIC_API_USE_32BIT_IDS (1ULL << 0) -#define KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK (1ULL << 1) +#define KVM_X2APIC_API_USE_32BIT_IDS (1ULL << 0) +#define KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK (1ULL << 1) +#define KVM_X2APIC_API_DISABLE_IGNORE_SUPPRESS_EOI_BROADCAST_QUIRK (1ULL << 2) +#define KVM_X2APIC_API_DISABLE_SUPPRESS_EOI_BROADCAST (1ULL << 3) struct kvm_hyperv_eventfd { __u32 conn_id; diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c index 0ae7f913d782..cf8a2162872b 100644 --- a/arch/x86/kvm/lapic.c +++ b/arch/x86/kvm/lapic.c @@ -562,6 +562,7 @@ void kvm_apic_set_version(struct kvm_vcpu *vcpu) * IOAPIC. */ if (guest_cpu_cap_has(vcpu, X86_FEATURE_X2APIC) && + !vcpu->kvm->arch.x2apic_disable_suppress_eoi_broadcast && !ioapic_in_kernel(vcpu->kvm)) v |= APIC_LVR_DIRECTED_EOI; kvm_lapic_set_reg(apic, APIC_LVR, v); @@ -1517,6 +1518,18 @@ static void kvm_ioapic_send_eoi(struct kvm_lapic *apic, int vector) /* Request a KVM exit to inform the userspace IOAPIC. */ if (irqchip_split(apic->vcpu->kvm)) { + /* + * Don't exit to userspace if the guest has enabled Directed + * EOI, a.k.a. Suppress EOI Broadcasts, in which case the local + * APIC doesn't broadcast EOIs (the guest must EOI the target + * I/O APIC(s) directly). Ignore the suppression if userspace + * has NOT disabled KVM's quirk (KVM advertised support for + * Suppress EOI Broadcasts without actually suppressing EOIs). + */ + if ((kvm_lapic_get_reg(apic, APIC_SPIV) & APIC_SPIV_DIRECTED_EOI) && + apic->vcpu->kvm->arch.disable_ignore_suppress_eoi_broadcast_quirk) + return; + apic->vcpu->arch.pending_ioapic_eoi = vector; kvm_make_request(KVM_REQ_IOAPIC_EOI_EXIT, apic->vcpu); return; diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index c9c2aa6f4705..e1b6fe783615 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -121,8 +121,11 @@ static u64 __read_mostly efer_reserved_bits = ~((u64)EFER_SCE); #define KVM_CAP_PMU_VALID_MASK KVM_PMU_CAP_DISABLE -#define KVM_X2APIC_API_VALID_FLAGS (KVM_X2APIC_API_USE_32BIT_IDS | \ - KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK) +#define KVM_X2APIC_API_VALID_FLAGS \ + (KVM_X2APIC_API_USE_32BIT_IDS | \ + KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK | \ + KVM_X2APIC_API_DISABLE_IGNORE_SUPPRESS_EOI_BROADCAST_QUIRK | \ + KVM_X2APIC_API_DISABLE_SUPPRESS_EOI_BROADCAST) static void update_cr8_intercept(struct kvm_vcpu *vcpu); static void process_nmi(struct kvm_vcpu *vcpu); @@ -6782,7 +6785,10 @@ int kvm_vm_ioctl_enable_cap(struct kvm *kvm, kvm->arch.x2apic_format = true; if (cap->args[0] & KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK) kvm->arch.x2apic_broadcast_quirk_disabled = true; - + if (cap->args[0] & KVM_X2APIC_API_DISABLE_IGNORE_SUPPRESS_EOI_BROADCAST_QUIRK) + kvm->arch.disable_ignore_suppress_eoi_broadcast_quirk = true; + if (cap->args[0] & KVM_X2APIC_API_DISABLE_SUPPRESS_EOI_BROADCAST) + kvm->arch.x2apic_disable_suppress_eoi_broadcast = true; r = 0; break; case KVM_CAP_X86_DISABLE_EXITS: -- 2.39.3

2 weeks, 5 days

5
19
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror