Fixes and cleanup for DRM fbdev emulation and deferred I/O
Subsystems (5)
v2 cover Fixes and cleanup for DRM fbdev emulation and deferred I/O
Hello,
This patch series contains two fixes and a cleanup for things that I noticed while debugging a regression in the fbdev emulation for a DRM driver.
The first two patches are trivial and shoulnd't be controversial, the third patch is less trivial, but it has been already reviewed by Thomas and I did test it to make sure that works as expected. With it, I got rid of the WARN that happened due a mutex used after it has been destroyed.
Best regards,
Javier
Changes in v2:
- Re-introduce the CONFIG_FB_DEFERRED_IO ifdef guard for the @fbdefio field declaration since the kernel test robot reported that's needed at the end.
Javier Martinez Canillas (3):
fbdev: Remove unused struct fb_deferred_io .first_io field drm/fb-helper: Check fb_deferred_io_init() return value drm/fb-helper: Use a per-driver FB deferred I/O handler
drivers/gpu/drm/drm_fbdev_generic.c | 15 ++++++++------- drivers/video/fbdev/core/fb_defio.c | 4 ---- include/drm/drm_fb_helper.h | 12 ++++++++++++ include/linux/fb.h | 1 - 4 files changed, 20 insertions(+), 12 deletions(-)
Signature · 1 line
2.39.0
PATCH v2 1/3 fbdev: Remove unused struct fb_deferred_io .first_io field
This optional callback was added in the commit 1f45f9dbb392 ("fb_defio:
add first_io callback") but it was never used by a driver. Let's remove it since it's unlikely that will be used after a decade that was added.
---
(no changes since v1)
drivers/video/fbdev/core/fb_defio.c | 4 ---- include/linux/fb.h | 1 - 2 files changed, 5 deletions(-)
drivers/video/fbdev/core/fb_defio.c +0 -4
@@ -157,10 +157,6 @@ static vm_fault_t fb_deferred_io_track_page(struct fb_info *info, unsigned long/* protect against the workqueue changing the page list */ mutex_lock (&fbdefio->lock); -/* first write in this cycle, notify the driver */ -if (fbdefio->first_io &&list_empty (&fbdefio->pagereflist)) - fbdefio->first_io (info); - pageref =fb_deferred_io_pageref_get (info, offset, page);if (WARN_ON_ONCE (!pageref)) { ret =VM_FAULT_OOM ;
include/linux/fb.h +0 -1
@@ -215,7 +215,6 @@ struct fb_deferred_io {struct mutex lock;/* mutex that protects the pageref list */ struct list_head pagereflist;/* list of pagerefs for touched pages */ /* callback */ -void (*first_io)(struct fb_info *info); void (*deferred_io)(struct fb_info *info, struct list_head *pagelist); };#endif
Signature · 1 line
2.39.0
Re: [PATCH v2 1/3] fbdev: Remove unused struct fb_deferred_io .first_io field
Quoted 7 lines · Javier Martinez Canillas
This optional callback was added in the commit
1f45f9dbb392("fb_defio:
addfirst_iocallback") but it was never used by a driver. Let's remove it since it's unlikely that will be used after a decade that was added.Signed-off-by Javier Martinez Canillas <[email protected]>Reviewed-by Thomas Zimmermann <[email protected]>---
Pushed this to drm-misc (drm-misc-next). Thanks!
Signature · 5 lines
Best regards, Javier Martinez Canillas Core Platforms Red Hat
PATCH v2 2/3 drm/fb-helper: Check fb_deferred_io_init() return value
In reply to Javier Martinez CanillasThe fb_deferred_io_init() can fail and return an errno code but currently there is no check for its return value.
Fix that and propagate to errno to the caller in the case of a failure.
---
(no changes since v1)
drivers/gpu/drm/drm_fbdev_generic.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/gpu/drm/drm_fbdev_generic.c +3 -1
@@ -223,7 +223,9 @@ static int drm_fbdev_fb_probe(struct drm_fb_helper *fb_helper, fbi->flags |=FBINFO_VIRTFB |FBINFO_READS_FAST ; fbi->fbdefio = &drm_fbdev_defio; -fb_deferred_io_init (fbi); + ret =fb_deferred_io_init (fbi); +if (ret) +return ret; }else {/* buffer is mapped for HW framebuffer */ ret =drm_client_buffer_vmap (fb_helper->buffer, &map);
Signature · 1 line
2.39.0
Re: [PATCH v2 2/3] drm/fb-helper: Check fb_deferred_io_init() return value
Quoted 9 lines · Javier Martinez Canillas
The
fb_deferred_io_init()can fail and return an errno code but currently there is no check for its return value.Fix that and propagate to errno to the caller in the case of a failure.
Fixes d536540f304c ("drm/fb-helper: Add generic fbdev emulation .fb_probe function")Signed-off-by Javier Martinez Canillas <[email protected]>Reviewed-by Thomas Zimmermann <[email protected]>---
Pushed this to drm-misc (drm-misc-fixes). Thanks!
Signature · 5 lines
Best regards, Javier Martinez Canillas Core Platforms Red Hat
PATCH v2 3/3 drm/fb-helper: Use a per-driver FB deferred I/O handler
In reply to Javier Martinez CanillasThe DRM fbdev emulation layer sets the struct fb_info .fbdefio field to a struct fb_deferred_io pointer, that is shared across all drivers that use the generic drm_fbdev_generic_setup() helper function.
It is a problem because the fbdev core deferred I/O logic assumes that the struct fb_deferred_io data is not shared between devices, and it's stored there state such as the list of pages touched and a mutex that is use to synchronize between the fb_deferred_io_track_page() function that track the dirty pages and fb_deferred_io_work() workqueue handler doing the actual deferred I/O.
The latter can lead to the following error, since it may happen that two drivers are probed and then one is removed, which causes the mutex bo be destroyed and not existing anymore by the time the other driver tries to grab it for the fbdev deferred I/O logic:
[ 369.756553] ------------[ cut here ]------------ [ 369.756604] DEBUG_LOCKS_WARN_ON(lock->magic != lock) [ 369.756631] WARNING: CPU: 2 PID: 1023 at kernel/locking/mutex.c:582 __mutex_lock+0x348/0x424 [ 369.756744] Modules linked in: nf_conntrack_netbios_ns nf_conntrack_broadcast nft_fib_inet nft_fib_ipv4 nft_fib_ipv6 nft_fib nft_reject_inet nf_reject_ipv4 nf_reject_ipv6 nft_reject nft_ct nft_chain_nat nf_nat nf_conntrack nf_defrag_ip v6 nf_defrag_ipv4 ip_set nf_tables nfnetlink qrtr btsdio bluetooth sunrpc brcmfmac snd_soc_hdmi_codec cpufreq_dt cfg80211 vfat fat vc4 rfkill brcmutil raspberrypi_cpufreq i2c_bcm2835 iproc_rng200 bcm2711_thermal snd_soc_core snd_pcm_dmaen gine leds_gpio nvmem_rmem joydev hid_cherry uas usb_storage gpio_raspberrypi_exp v3d snd_pcm raspberrypi_hwmon gpu_sched bcm2835_wdt broadcom bcm_phy_lib snd_timer genet snd mdio_bcm_unimac clk_bcm2711_dvp soundcore drm_display_helper pci e_brcmstb cec ip6_tables ip_tables fuse [ 369.757400] CPU: 2 PID: 1023 Comm: fbtest Not tainted 5.19.0-rc6+ #94 [ 369.757455] Hardware name: raspberrypi,4-model-b Raspberry Pi 4 Model B Rev 1.4/Raspberry Pi 4 Model B Rev 1.4, BIOS 2022.10 10/01/2022 [ 369.757538] pstate: 00400005 (nzcv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--) [ 369.757596] pc : __mutex_lock+0x348/0x424 [ 369.757635] lr : __mutex_lock+0x348/0x424 [ 369.757672] sp : ffff80000953bb00 [ 369.757703] x29: ffff80000953bb00 x28: ffff17fdc087c000 x27: 0000000000000002 [ 369.757771] x26: ffff17fdc349f9b0 x25: fffffc5ff72e0100 x24: 0000000000000000 [ 369.757838] x23: 0000000000000000 x22: 0000000000000002 x21: ffffa618df636f10 [ 369.757903] x20: ffff80000953bb68 x19: ffffa618e0f18138 x18: 0000000000000001 [ 369.757968] x17: 0000000020000000 x16: 0000000000000002 x15: 0000000000000000 [ 369.758032] x14: 0000000000000000 x13: 284e4f5f4e524157 x12: 5f534b434f4c5f47 [ 369.758097] x11: 00000000ffffdfff x10: ffffa618e0c79f88 x9 : ffffa618de472484 [ 369.758162] x8 : 000000000002ffe8 x7 : c0000000ffffdfff x6 : 00000000000affa8 [ 369.758227] x5 : 0000000000001fff x4 : 0000000000000000 x3 : 0000000000000027 [ 369.758292] x2 : 0000000000000001 x1 : ffff17fdc087c000 x0 : 0000000000000028 [ 369.758357] Call trace: [ 369.758383] __mutex_lock+0x348/0x424 [ 369.758420] mutex_lock_nested+0x4c/0x5c [ 369.758459] fb_deferred_io_mkwrite+0x78/0x1d8 [ 369.758507] do_page_mkwrite+0x5c/0x19c [ 369.758550] wp_page_shared+0x70/0x1a0 [ 369.758590] do_wp_page+0x3d0/0x510 [ 369.758628] handle_pte_fault+0x1c0/0x1e0 [ 369.758670] __handle_mm_fault+0x250/0x380 [ 369.758712] handle_mm_fault+0x17c/0x3a4 [ 369.758753] do_page_fault+0x158/0x530 [ 369.758792] do_mem_abort+0x50/0xa0 [ 369.758831] el0_da+0x78/0x19c [ 369.758864] el0t_64_sync_handler+0xbc/0x150 [ 369.758904] el0t_64_sync+0x190/0x194 [ 369.758942] irq event stamp: 11395 [ 369.758973] hardirqs last enabled at (11395): [<ffffa618de472554>] __up_console_sem+0x74/0x80 [ 369.759042] hardirqs last disabled at (11394): [<ffffa618de47254c>] __up_console_sem+0x6c/0x80 [ 369.760554] softirqs last enabled at (11392): [<ffffa618de330a74>] __do_softirq+0x4c4/0x6b8 [ 369.762060] softirqs last disabled at (11383): [<ffffa618de3c9124>] __irq_exit_rcu+0x104/0x214 [ 369.763564] ---[ end trace 0000000000000000 ]---
---
Changes in v2:
- Re-introduce the CONFIG_FB_DEFERRED_IO ifdef guard for the @fbdefio field declaration since the kernel test robot reported that's needed at the end.
drivers/gpu/drm/drm_fbdev_generic.c | 11 +++++------ include/drm/drm_fb_helper.h | 12 ++++++++++++ 2 files changed, 17 insertions(+), 6 deletions(-)
drivers/gpu/drm/drm_fbdev_generic.c +5 -6
@@ -171,11 +171,6 @@ static const struct fb_ops drm_fbdev_fb_ops = { .fb_imageblit = drm_fbdev_fb_imageblit, }; -static struct fb_deferred_io drm_fbdev_defio = { - .delay =HZ /20 , - .deferred_io = drm_fb_helper_deferred_io, -}; -/* * This function uses the client API to create a framebuffer backed by a dumb buffer. */
@@ -222,7 +217,11 @@ static int drm_fbdev_fb_probe(struct drm_fb_helper *fb_helper,return -ENOMEM ; fbi->flags |=FBINFO_VIRTFB |FBINFO_READS_FAST ; - fbi->fbdefio = &drm_fbdev_defio; +/* Set a default deferred I/O handler */ + fb_helper->fbdefio.delay =HZ /20 ; + fb_helper->fbdefio.deferred_io = drm_fb_helper_deferred_io; + + fbi->fbdefio = &fb_helper->fbdefio; ret =fb_deferred_io_init (fbi);if (ret)return ret;
include/drm/drm_fb_helper.h +12 -0
@@ -203,6 +203,18 @@ struct drm_fb_helper {* the smem_start field should always be cleared to zero. */ bool hint_leak_smem_start; + +#ifdef CONFIG_FB_DEFERRED_IO +/** +* @fbdefio: +* +* Temporary storage for the driver's FB deferred I/O handler. If the +* driver uses the DRM fbdev emulation layer, this is set by the core +* to a generic deferred I/O handler if a driver is preferring to use +* a shadow buffer. +*/ +struct fb_deferred_io fbdefio; +#endif }; static inline struct drm_fb_helper *
Signature · 1 line
2.39.0
Re: [PATCH v2 3/3] drm/fb-helper: Use a per-driver FB deferred I/O handler
Quoted 4 lines · Javier Martinez Canillas
The DRM fbdev emulation layer sets the struct
fb_info.fbdefio field to a structfb_deferred_iopointer, that is shared across all drivers that use the genericdrm_fbdev_generic_setup()helper function.
[...]
Quoted 6 lines
Fixes d536540f304c ("drm/fb-helper: Add generic fbdev emulation .fb_probe function")Signed-off-by Javier Martinez Canillas <[email protected]>Reviewed-by Thomas Zimmermann <[email protected]>---
Pushed this to drm-misc (drm-misc-fixes). Thanks!
Signature · 5 lines
Best regards, Javier Martinez Canillas Core Platforms Red Hat