[PATCH] usb: mtu3: fix panic in mtu3_gadget_disconnect()

List overview All Threads
Download

newer

older

[PATCH] crypto: af_alg - avoid...

[PATCH AUTOSEL 4.19 01/60]...

Macpaul Lin

31 Jul 2020 31 Jul '20

6:36 a.m.

This patch fixes a possible issue when mtu3_gadget_stop() already assigned NULL to mtu->gadget_driver during mtu_gadget_disconnect().

Backtrace: [<ffffff9008161974>] notifier_call_chain+0xa4/0x128 [<ffffff9008161fd4>] __atomic_notifier_call_chain+0x84/0x138 [<ffffff9008162ec0>] notify_die+0xb0/0x120 [<ffffff900809e340>] die+0x1f8/0x5d0 [<ffffff90080d03b4>] __do_kernel_fault+0x19c/0x280 [<ffffff90080d04dc>] do_bad_area+0x44/0x140 [<ffffff90080d0f9c>] do_translation_fault+0x4c/0x90 [<ffffff9008080a78>] do_mem_abort+0xb8/0x258 [<ffffff90080849d0>] el1_da+0x24/0x3c [<ffffff9009bde01c>] mtu3_gadget_disconnect+0xac/0x128 [<ffffff9009bd576c>] mtu3_irq+0x34c/0xc18 [<ffffff90082ac03c>] __handle_irq_event_percpu+0x2ac/0xcd0 [<ffffff90082acae0>] handle_irq_event_percpu+0x80/0x138 [<ffffff90082acc44>] handle_irq_event+0xac/0x148 [<ffffff90082b71cc>] handle_fasteoi_irq+0x234/0x568 [<ffffff90082a8708>] generic_handle_irq+0x48/0x68 [<ffffff90082a96ac>] __handle_domain_irq+0x264/0x1740 [<ffffff90080819f4>] gic_handle_irq+0x14c/0x250 [<ffffff9008084cec>] el1_irq+0xec/0x194 [<ffffff90085b985c>] dma_pool_alloc+0x6e4/0xae0 [<ffffff9008d7f890>] cmdq_mbox_pool_alloc_impl+0xb0/0x238 [<ffffff9008d80904>] cmdq_pkt_alloc_buf+0x2dc/0x7c0 [<ffffff9008d80f60>] cmdq_pkt_add_cmd_buffer+0x178/0x270 [<ffffff9008d82320>] cmdq_pkt_perf_begin+0x108/0x148 [<ffffff9008d824d8>] cmdq_pkt_create+0x178/0x1f0 [<ffffff9008f96230>] mtk_crtc_config_default_path+0x328/0x7a0 [<ffffff90090246cc>] mtk_drm_idlemgr_kick+0xa6c/0x1460 [<ffffff9008f9bbb4>] mtk_drm_crtc_atomic_begin+0x1a4/0x1a68 [<ffffff9008e8df9c>] drm_atomic_helper_commit_planes+0x154/0x878 [<ffffff9008f2fb70>] mtk_atomic_complete.isra.16+0xe80/0x19c8 [<ffffff9008f30910>] mtk_atomic_commit+0x258/0x898 [<ffffff9008ef142c>] drm_atomic_commit+0xcc/0x108 [<ffffff9008ef7cf0>] drm_mode_atomic_ioctl+0x1c20/0x2580 [<ffffff9008ebc768>] drm_ioctl_kernel+0x118/0x1b0 [<ffffff9008ebcde8>] drm_ioctl+0x5c0/0x920 [<ffffff900863b030>] do_vfs_ioctl+0x188/0x1820 [<ffffff900863c754>] SyS_ioctl+0x8c/0xa0

Signed-off-by: Macpaul Lin macpaul.lin@mediatek.com Cc: stable@vger.kernel.org --- drivers/usb/mtu3/mtu3_gadget.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/drivers/usb/mtu3/mtu3_gadget.c b/drivers/usb/mtu3/mtu3_gadget.c index 68ea4395f871..f20fb83b3239 100644 --- a/drivers/usb/mtu3/mtu3_gadget.c +++ b/drivers/usb/mtu3/mtu3_gadget.c @@ -843,7 +843,12 @@ void mtu3_gadget_disconnect(struct mtu3 *mtu) dev_dbg(mtu->dev, "gadget DISCONNECT\n"); if (mtu->gadget_driver && mtu->gadget_driver->disconnect) { spin_unlock(&mtu->lock); - mtu->gadget_driver->disconnect(&mtu->g); + /* + * avoid kernel panic because mtu3_gadget_stop() assigned NULL + * to mtu->gadget_driver. + */ + if (mtu->gadget_driver && mtu->gadget_driver->disconnect) + mtu->gadget_driver->disconnect(&mtu->g); spin_lock(&mtu->lock); }

-- 2.18.0

Show replies by date

Macpaul Lin

31 Jul 31 Jul

8:57 a.m.

New subject: [PATCH v2] usb: mtu3: fix panic in mtu3_gadget_disconnect()

This patch fixes a possible issue when mtu3_gadget_stop() already assigned NULL to mtu->gadget_driver during mtu_gadget_disconnect().

[<ffffff9008161974>] notifier_call_chain+0xa4/0x128 [<ffffff9008161fd4>] __atomic_notifier_call_chain+0x84/0x138 [<ffffff9008162ec0>] notify_die+0xb0/0x120 [<ffffff900809e340>] die+0x1f8/0x5d0 [<ffffff90080d03b4>] __do_kernel_fault+0x19c/0x280 [<ffffff90080d04dc>] do_bad_area+0x44/0x140 [<ffffff90080d0f9c>] do_translation_fault+0x4c/0x90 [<ffffff9008080a78>] do_mem_abort+0xb8/0x258 [<ffffff90080849d0>] el1_da+0x24/0x3c [<ffffff9009bde01c>] mtu3_gadget_disconnect+0xac/0x128 [<ffffff9009bd576c>] mtu3_irq+0x34c/0xc18 [<ffffff90082ac03c>] __handle_irq_event_percpu+0x2ac/0xcd0 [<ffffff90082acae0>] handle_irq_event_percpu+0x80/0x138 [<ffffff90082acc44>] handle_irq_event+0xac/0x148 [<ffffff90082b71cc>] handle_fasteoi_irq+0x234/0x568 [<ffffff90082a8708>] generic_handle_irq+0x48/0x68 [<ffffff90082a96ac>] __handle_domain_irq+0x264/0x1740 [<ffffff90080819f4>] gic_handle_irq+0x14c/0x250 [<ffffff9008084cec>] el1_irq+0xec/0x194 [<ffffff90085b985c>] dma_pool_alloc+0x6e4/0xae0 [<ffffff9008d7f890>] cmdq_mbox_pool_alloc_impl+0xb0/0x238 [<ffffff9008d80904>] cmdq_pkt_alloc_buf+0x2dc/0x7c0 [<ffffff9008d80f60>] cmdq_pkt_add_cmd_buffer+0x178/0x270 [<ffffff9008d82320>] cmdq_pkt_perf_begin+0x108/0x148 [<ffffff9008d824d8>] cmdq_pkt_create+0x178/0x1f0 [<ffffff9008f96230>] mtk_crtc_config_default_path+0x328/0x7a0 [<ffffff90090246cc>] mtk_drm_idlemgr_kick+0xa6c/0x1460 [<ffffff9008f9bbb4>] mtk_drm_crtc_atomic_begin+0x1a4/0x1a68 [<ffffff9008e8df9c>] drm_atomic_helper_commit_planes+0x154/0x878 [<ffffff9008f2fb70>] mtk_atomic_complete.isra.16+0xe80/0x19c8 [<ffffff9008f30910>] mtk_atomic_commit+0x258/0x898 [<ffffff9008ef142c>] drm_atomic_commit+0xcc/0x108 [<ffffff9008ef7cf0>] drm_mode_atomic_ioctl+0x1c20/0x2580 [<ffffff9008ebc768>] drm_ioctl_kernel+0x118/0x1b0 [<ffffff9008ebcde8>] drm_ioctl+0x5c0/0x920 [<ffffff900863b030>] do_vfs_ioctl+0x188/0x1820 [<ffffff900863c754>] SyS_ioctl+0x8c/0xa0

Signed-off-by: Macpaul Lin macpaul.lin@mediatek.com Cc: stable@vger.kernel.org --- Changes for v2: - Check mtu_gadget_driver out of spin_lock might still not work. We use a temporary pointer to keep the callback function.

drivers/usb/mtu3/mtu3_gadget.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-)

diff --git a/drivers/usb/mtu3/mtu3_gadget.c b/drivers/usb/mtu3/mtu3_gadget.c index 68ea4395f871..40cb6626f496 100644 --- a/drivers/usb/mtu3/mtu3_gadget.c +++ b/drivers/usb/mtu3/mtu3_gadget.c @@ -840,10 +840,17 @@ void mtu3_gadget_suspend(struct mtu3 *mtu) /* called when VBUS drops below session threshold, and in other cases */ void mtu3_gadget_disconnect(struct mtu3 *mtu) { + struct usb_gadget_driver *driver; + dev_dbg(mtu->dev, "gadget DISCONNECT\n"); if (mtu->gadget_driver && mtu->gadget_driver->disconnect) { + driver = mtu->gadget_driver; spin_unlock(&mtu->lock); - mtu->gadget_driver->disconnect(&mtu->g); + /* + * avoid kernel panic because mtu3_gadget_stop() assigned NULL + * to mtu->gadget_driver. + */ + driver->disconnect(&mtu->g); spin_lock(&mtu->lock); }

-- 2.18.0

Alan Stern

2:22 p.m.

New subject: [PATCH v2] usb: mtu3: fix panic in mtu3_gadget_disconnect()

On Fri, Jul 31, 2020 at 04:57:58PM +0800, Macpaul Lin wrote:

...

This patch fixes a possible issue when mtu3_gadget_stop() already assigned NULL to mtu->gadget_driver during mtu_gadget_disconnect().

...

Signed-off-by: Macpaul Lin macpaul.lin@mediatek.com Cc: stable@vger.kernel.org

Changes for v2:

Check mtu_gadget_driver out of spin_lock might still not work. We use a temporary pointer to keep the callback function.

drivers/usb/mtu3/mtu3_gadget.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-)

diff --git a/drivers/usb/mtu3/mtu3_gadget.c b/drivers/usb/mtu3/mtu3_gadget.c index 68ea4395f871..40cb6626f496 100644 --- a/drivers/usb/mtu3/mtu3_gadget.c +++ b/drivers/usb/mtu3/mtu3_gadget.c @@ -840,10 +840,17 @@ void mtu3_gadget_suspend(struct mtu3 *mtu) /* called when VBUS drops below session threshold, and in other cases */ void mtu3_gadget_disconnect(struct mtu3 *mtu) {
struct usb_gadget_driver *driver;

dev_dbg(mtu->dev, "gadget DISCONNECT\n"); if (mtu->gadget_driver && mtu->gadget_driver->disconnect) {
driver = mtu->gadget_driver;
spin_unlock(&mtu->lock);
mtu->gadget_driver->disconnect(&mtu->g);
/*
 * avoid kernel panic because mtu3_gadget_stop() assigned NULL
 * to mtu->gadget_driver.
 */
driver->disconnect(&mtu->g);
spin_lock(&mtu->lock); }

This is not the right approach; it might race with the gadget driver unregistering itself.

Instead, mtu3_gadget_stop() should call synchronize_irq() after releasing the IRQ line. When synchronize_irq() returns, you'll know any IRQ handlers have finished running, so you won't receive any more disconnect notifications. Then it will be safe to acquire the spinlock and set mtu->gadget_driver to NULL.

Alan Stern

Macpaul Lin

27 Aug 27 Aug

9:22 a.m.

New subject: [PATCH v3] usb: mtu3: fix panic in mtu3_gadget_stop()

This patch fixes a possible issue when mtu3_gadget_stop() already assigned NULL to mtu->gadget_driver during mtu_gadget_disconnect().

Signed-off-by: Macpaul Lin macpaul.lin@mediatek.com Cc: stable@vger.kernel.org --- Changes for v3: - Call synchronize_irq() in mtu3_gadget_stop() instead of remembering callback function in mtu3_gadget_disconnect(). Thanks for Alan's suggestion.

Changes for v2: - Check mtu_gadget_driver out of spin_lock might still not work. We use a temporary pointer to remember the callback function.

drivers/usb/mtu3/mtu3_gadget.c | 1 + 1 file changed, 1 insertions(+)

diff --git a/drivers/usb/mtu3/mtu3_gadget.c b/drivers/usb/mtu3/mtu3_gadget.c index 1de5c9a..1ab3d3a 100644 --- a/drivers/usb/mtu3/mtu3_gadget.c +++ b/drivers/usb/mtu3/mtu3_gadget.c @@ -564,6 +564,7 @@ static int mtu3_gadget_stop(struct usb_gadget *g)

spin_unlock_irqrestore(&mtu->lock, flags);

+ synchronize_irq(mtu->irq); return 0; }

-- 1.7.9.5

Felipe Balbi

1:03 p.m.

New subject: [PATCH v3] usb: mtu3: fix panic in mtu3_gadget_stop()

Macpaul Lin macpaul.lin@mediatek.com writes:

...

This patch fixes a possible issue when mtu3_gadget_stop() already assigned NULL to mtu->gadget_driver during mtu_gadget_disconnect().

[<ffffff9008161974>] notifier_call_chain+0xa4/0x128 [<ffffff9008161fd4>] __atomic_notifier_call_chain+0x84/0x138 [<ffffff9008162ec0>] notify_die+0xb0/0x120 [<ffffff900809e340>] die+0x1f8/0x5d0 [<ffffff90080d03b4>] __do_kernel_fault+0x19c/0x280 [<ffffff90080d04dc>] do_bad_area+0x44/0x140 [<ffffff90080d0f9c>] do_translation_fault+0x4c/0x90 [<ffffff9008080a78>] do_mem_abort+0xb8/0x258 [<ffffff90080849d0>] el1_da+0x24/0x3c [<ffffff9009bde01c>] mtu3_gadget_disconnect+0xac/0x128 [<ffffff9009bd576c>] mtu3_irq+0x34c/0xc18 [<ffffff90082ac03c>] __handle_irq_event_percpu+0x2ac/0xcd0 [<ffffff90082acae0>] handle_irq_event_percpu+0x80/0x138 [<ffffff90082acc44>] handle_irq_event+0xac/0x148 [<ffffff90082b71cc>] handle_fasteoi_irq+0x234/0x568 [<ffffff90082a8708>] generic_handle_irq+0x48/0x68 [<ffffff90082a96ac>] __handle_domain_irq+0x264/0x1740 [<ffffff90080819f4>] gic_handle_irq+0x14c/0x250 [<ffffff9008084cec>] el1_irq+0xec/0x194 [<ffffff90085b985c>] dma_pool_alloc+0x6e4/0xae0 [<ffffff9008d7f890>] cmdq_mbox_pool_alloc_impl+0xb0/0x238 [<ffffff9008d80904>] cmdq_pkt_alloc_buf+0x2dc/0x7c0 [<ffffff9008d80f60>] cmdq_pkt_add_cmd_buffer+0x178/0x270 [<ffffff9008d82320>] cmdq_pkt_perf_begin+0x108/0x148 [<ffffff9008d824d8>] cmdq_pkt_create+0x178/0x1f0 [<ffffff9008f96230>] mtk_crtc_config_default_path+0x328/0x7a0 [<ffffff90090246cc>] mtk_drm_idlemgr_kick+0xa6c/0x1460 [<ffffff9008f9bbb4>] mtk_drm_crtc_atomic_begin+0x1a4/0x1a68 [<ffffff9008e8df9c>] drm_atomic_helper_commit_planes+0x154/0x878 [<ffffff9008f2fb70>] mtk_atomic_complete.isra.16+0xe80/0x19c8 [<ffffff9008f30910>] mtk_atomic_commit+0x258/0x898 [<ffffff9008ef142c>] drm_atomic_commit+0xcc/0x108 [<ffffff9008ef7cf0>] drm_mode_atomic_ioctl+0x1c20/0x2580 [<ffffff9008ebc768>] drm_ioctl_kernel+0x118/0x1b0 [<ffffff9008ebcde8>] drm_ioctl+0x5c0/0x920 [<ffffff900863b030>] do_vfs_ioctl+0x188/0x1820 [<ffffff900863c754>] SyS_ioctl+0x8c/0xa0

Signed-off-by: Macpaul Lin macpaul.lin@mediatek.com Cc: stable@vger.kernel.org

missing a Fixes: line here

-- balbi

Macpaul Lin

2:42 p.m.

New subject: [PATCH v4] usb: mtu3: fix panic in mtu3_gadget_stop()

This patch fixes a possible issue when mtu3_gadget_stop() already assigned NULL to mtu->gadget_driver during mtu_gadget_disconnect().

Fixes: df2069acb005 ("usb: Add MediaTek USB3 DRD driver") Signed-off-by: Macpaul Lin macpaul.lin@mediatek.com Cc: stable@vger.kernel.org --- Changes for v4: - Add a "Fixes:" line. Thanks Felipe. Changes for v3: - Call synchronize_irq() in mtu3_gadget_stop() instead of remembering callback function in mtu3_gadget_disconnect(). Thanks for Alan's suggestion. Changes for v2: - Check mtu_gadget_driver out of spin_lock might still not work. We use a temporary pointer to remember the callback function.

drivers/usb/mtu3/mtu3_gadget.c | 1 + 1 file changed, 1 insertions(+)

spin_unlock_irqrestore(&mtu->lock, flags);

+ synchronize_irq(mtu->irq); return 0; }

-- 1.7.9.5

Chunfeng Yun

31 Aug 31 Aug

1:50 a.m.

New subject: [PATCH v4] usb: mtu3: fix panic in mtu3_gadget_stop()

On Thu, 2020-08-27 at 22:42 +0800, Macpaul Lin wrote:

...

This patch fixes a possible issue when mtu3_gadget_stop() already assigned NULL to mtu->gadget_driver during mtu_gadget_disconnect().

[<ffffff9008161974>] notifier_call_chain+0xa4/0x128 [<ffffff9008161fd4>] __atomic_notifier_call_chain+0x84/0x138 [<ffffff9008162ec0>] notify_die+0xb0/0x120 [<ffffff900809e340>] die+0x1f8/0x5d0 [<ffffff90080d03b4>] __do_kernel_fault+0x19c/0x280 [<ffffff90080d04dc>] do_bad_area+0x44/0x140 [<ffffff90080d0f9c>] do_translation_fault+0x4c/0x90 [<ffffff9008080a78>] do_mem_abort+0xb8/0x258 [<ffffff90080849d0>] el1_da+0x24/0x3c [<ffffff9009bde01c>] mtu3_gadget_disconnect+0xac/0x128 [<ffffff9009bd576c>] mtu3_irq+0x34c/0xc18 [<ffffff90082ac03c>] __handle_irq_event_percpu+0x2ac/0xcd0 [<ffffff90082acae0>] handle_irq_event_percpu+0x80/0x138 [<ffffff90082acc44>] handle_irq_event+0xac/0x148 [<ffffff90082b71cc>] handle_fasteoi_irq+0x234/0x568 [<ffffff90082a8708>] generic_handle_irq+0x48/0x68 [<ffffff90082a96ac>] __handle_domain_irq+0x264/0x1740 [<ffffff90080819f4>] gic_handle_irq+0x14c/0x250 [<ffffff9008084cec>] el1_irq+0xec/0x194 [<ffffff90085b985c>] dma_pool_alloc+0x6e4/0xae0 [<ffffff9008d7f890>] cmdq_mbox_pool_alloc_impl+0xb0/0x238 [<ffffff9008d80904>] cmdq_pkt_alloc_buf+0x2dc/0x7c0 [<ffffff9008d80f60>] cmdq_pkt_add_cmd_buffer+0x178/0x270 [<ffffff9008d82320>] cmdq_pkt_perf_begin+0x108/0x148 [<ffffff9008d824d8>] cmdq_pkt_create+0x178/0x1f0 [<ffffff9008f96230>] mtk_crtc_config_default_path+0x328/0x7a0 [<ffffff90090246cc>] mtk_drm_idlemgr_kick+0xa6c/0x1460 [<ffffff9008f9bbb4>] mtk_drm_crtc_atomic_begin+0x1a4/0x1a68 [<ffffff9008e8df9c>] drm_atomic_helper_commit_planes+0x154/0x878 [<ffffff9008f2fb70>] mtk_atomic_complete.isra.16+0xe80/0x19c8 [<ffffff9008f30910>] mtk_atomic_commit+0x258/0x898 [<ffffff9008ef142c>] drm_atomic_commit+0xcc/0x108 [<ffffff9008ef7cf0>] drm_mode_atomic_ioctl+0x1c20/0x2580 [<ffffff9008ebc768>] drm_ioctl_kernel+0x118/0x1b0 [<ffffff9008ebcde8>] drm_ioctl+0x5c0/0x920 [<ffffff900863b030>] do_vfs_ioctl+0x188/0x1820 [<ffffff900863c754>] SyS_ioctl+0x8c/0xa0

Fixes: df2069acb005 ("usb: Add MediaTek USB3 DRD driver") Signed-off-by: Macpaul Lin macpaul.lin@mediatek.com Cc: stable@vger.kernel.org

Changes for v4:

Add a "Fixes:" line. Thanks Felipe.

Changes for v3:

Call synchronize_irq() in mtu3_gadget_stop() instead of remembering callback function in mtu3_gadget_disconnect(). Thanks for Alan's suggestion.

Changes for v2:

Check mtu_gadget_driver out of spin_lock might still not work. We use a temporary pointer to remember the callback function.

drivers/usb/mtu3/mtu3_gadget.c | 1 + 1 file changed, 1 insertions(+)

diff --git a/drivers/usb/mtu3/mtu3_gadget.c b/drivers/usb/mtu3/mtu3_gadget.c index 1de5c9a..1ab3d3a 100644 --- a/drivers/usb/mtu3/mtu3_gadget.c +++ b/drivers/usb/mtu3/mtu3_gadget.c @@ -564,6 +564,7 @@ static int mtu3_gadget_stop(struct usb_gadget *g) spin_unlock_irqrestore(&mtu->lock, flags);

synchronize_irq(mtu->irq); return 0;

}

Acked-by: Chunfeng Yun chunfeng.yun@mediatek.com

Thanks a lot

Macpaul Lin

6 Nov 6 Nov

5:54 a.m.

New subject: [RESEND PATCH v4] usb: mtu3: fix panic in mtu3_gadget_stop()

This patch fixes a possible issue when mtu3_gadget_stop() already assigned NULL to mtu->gadget_driver during mtu_gadget_disconnect().

Fixes: df2069acb005 ("usb: Add MediaTek USB3 DRD driver") Signed-off-by: Macpaul Lin macpaul.lin@mediatek.com Acked-by: Chunfeng Yun chunfeng.yun@mediatek.com Cc: stable@vger.kernel.org --- RESEND for v4: - Resend this patch by plain-text instead of MTK IT's default (base64) outgoing SMTP settings. - Add Acked-by: Chunfeng Yun chunfeng.yun@mediatek.com Changes for v4: - Add a "Fixes:" line. Thanks Felipe. Changes for v3: - Call synchronize_irq() in mtu3_gadget_stop() instead of remembering callback function in mtu3_gadget_disconnect(). Thanks for Alan's suggestion. Changes for v2: - Check mtu_gadget_driver out of spin_lock might still not work. We use a temporary pointer to remember the callback function.

drivers/usb/mtu3/mtu3_gadget.c | 1 + 1 file changed, 1 insertions(+)

spin_unlock_irqrestore(&mtu->lock, flags);

+ synchronize_irq(mtu->irq); return 0; }

-- 1.7.9.5

1278

days inactive

1376

days old

linux-stable-mirror@lists.linaro.org

7 comments

participants

tags (0)

participants (4)

Alan Stern
Chunfeng Yun
Felipe Balbi
Macpaul Lin