Message-ID: <5500a71a-68bd-4dd1-99cb-6523281f0c7d@amd.com>
Date: Mon, 11 Aug 2025 17:05:39 +0800
From: "Du, Bin" <bin.du@....com>
To: Sultan Alsawaf <sultan@...neltoast.com>
Cc: mchehab@...nel.org, hverkuil@...all.nl,
 laurent.pinchart+renesas@...asonboard.com, bryan.odonoghue@...aro.org,
 sakari.ailus@...ux.intel.com, prabhakar.mahadev-lad.rj@...renesas.com,
 linux-media@...r.kernel.org, linux-kernel@...r.kernel.org,
 pratap.nirujogi@....com, benjamin.chan@....com, king.li@....com,
 gjorgji.rosikopulos@....com, Phil.Jawich@....com, Dominic.Antony@....com
Subject: Re: [PATCH v2 6/8] media: platform: amd: isp4 video node and buffers
 handling added

Thanks Sultan.

On 8/11/2025 2:02 PM, Sultan Alsawaf wrote:
> On Wed, Jul 30, 2025 at 05:34:22PM -0700, Sultan Alsawaf wrote:
>> On Tue, Jul 29, 2025 at 03:43:14PM +0800, Du, Bin wrote:
>>> Hi Sultan, really appreciate your time and effort
>>>
>>> On 7/28/2025 3:04 PM, Sultan Alsawaf wrote:
>>>> I found more refcounting issues in addition to the ones from my other emails
>>>> while trying to make my webcam work:
>>>>
>>>> On Wed, Jun 18, 2025 at 05:19:57PM +0800, Bin Du wrote:
>>>>> +static int isp4vid_vb2_mmap(void *buf_priv, struct vm_area_struct *vma)
>>>>> +{
>>>>> +	struct isp4vid_vb2_buf *buf = buf_priv;
>>>>> +	int ret;
>>>>> +
>>>>> +	if (!buf) {
>>>>> +		pr_err("fail no memory to map\n");
>>>>> +		return -EINVAL;
>>>>> +	}
>>>>> +
>>>>> +	ret = remap_vmalloc_range(vma, buf->vaddr, 0);
>>>>> +	if (ret) {
>>>>> +		dev_err(buf->dev, "fail remap vmalloc mem, %d\n", ret);
>>>>> +		return ret;
>>>>> +	}
>>>>> +
>>>>> +	/*
>>>>> +	 * Make sure that vm_areas for 2 buffers won't be merged together
>>>>> +	 */
>>>>> +	vm_flags_set(vma, VM_DONTEXPAND);
>>>>> +
>>>>> +	dev_dbg(buf->dev, "mmap isp user bo 0x%llx size %ld refcount %d\n",
>>>>> +		buf->gpu_addr, buf->size, buf->refcount.refs.counter);
>>>>
>>>> Use refcount_read() instead of reading the refcount's atomic_t counter directly.
>>>> This is done in 3 other places; change those to refcount_read() as well.
>>>>
>>>> This didn't cause any functional problems, but it should still be fixed.
>>>>
>>>>> +
>>>>> +	return 0;
>>>>> +}
>>>>
>>> Sure, will fix it
>>>
>>>> [snip]
>>>>
>>>>> +static void isp4vid_vb2_detach_dmabuf(void *mem_priv)
>>>>> +{
>>>>> +	struct isp4vid_vb2_buf *buf = mem_priv;
>>>>> +
>>>>> +	if (!buf) {
>>>>> +		pr_err("fail invalid buf handle\n");
>>>>> +		return;
>>>>> +	}
>>>>> +
>>>>> +	struct iosys_map map = IOSYS_MAP_INIT_VADDR(buf->vaddr);
>>>>> +
>>>>> +	dev_dbg(buf->dev, "detach dmabuf of isp user bo 0x%llx size %ld",
>>>>> +		buf->gpu_addr, buf->size);
>>>>> +
>>>>> +	if (buf->vaddr)
>>>>> +		dma_buf_vunmap_unlocked(buf->dbuf, &map);
>>>>> +
>>>>> +	// put dmabuf for exported ones
>>>>> +	dma_buf_put(buf->dbuf);
>>>>> +
>>>>> +	kfree(buf);
>>>>> +}
>>>>
>>>> As mentioned in the other email, the dma_buf_put() here needs to be removed. But
>>>> that's not all: the dma_buf_vunmap_unlocked() needs to be removed too because
>>>> vb2 will always unmap the buffer before detaching it. As a result, having the
>>>> dma_buf_vunmap_unlocked() call here results in a use-after-free when vb2 calls
>>>> the unmap_dmabuf memop.
>>>>
>>>> Change this function to the following:
>>>>
>>>> 	static void isp4vid_vb2_detach_dmabuf(void *mem_priv)
>>>> 	{
>>>> 		struct isp4vid_vb2_buf *buf = mem_priv;
>>>>
>>>> 		kfree(buf);
>>>> 	}
>>>>
>>> Will have a try
>>>
>>>>> +static void isp4vid_qops_buffer_cleanup(struct vb2_buffer *vb)
>>>>> +{
>>>>> +	struct isp4vid_dev *isp_vdev = vb2_get_drv_priv(vb->vb2_queue);
>>>>> +	struct isp4vid_vb2_buf *buf = vb->planes[0].mem_priv;
>>>>> +
>>>>> +	dev_dbg(isp_vdev->dev, "%s|index=%u vb->memory %u",
>>>>> +		isp_vdev->vdev.name, vb->index, vb->memory);
>>>>> +
>>>>> +	if (!buf) {
>>>>> +		dev_err(isp_vdev->dev, "Invalid buf handle");
>>>>> +		return;
>>>>> +	}
>>>>> +
>>>>> +	// release implicit dmabuf reference here for vb2 buffer
>>>>> +	// of type MMAP and is exported
>>>>> +	if (vb->memory == VB2_MEMORY_MMAP && buf->is_expbuf) {
>>>>> +		dma_buf_put(buf->dbuf);
>>>>> +		dev_dbg(isp_vdev->dev,
>>>>> +			"put dmabuf for vb->memory %d expbuf %d",
>>>>> +			vb->memory,
>>>>> +			buf->is_expbuf);
>>>>> +	}
>>>>> +}
>>>>> +
>>>>
>>>> Remove the isp4vid_qops_buffer_cleanup() function. It causes a use-after-free by
>>>> doing an extra dma_buf_put(). This function isn't needed now that the refcount
>>>> issues are solved.
>>>>
>>> Will have a try
>>>
>>>> [snip]
>>>>
>>>>> +static const struct vb2_ops isp4vid_qops = {
>>>>> +	.queue_setup = isp4vid_qops_queue_setup,
>>>>> +	.buf_cleanup = isp4vid_qops_buffer_cleanup,
>>>>
>>>> Remove the .buf_cleanup hook too.
>>>>
>>> Will have a try
>>>
>>>>> +	.buf_queue = isp4vid_qops_buffer_queue,
>>>>> +	.start_streaming = isp4vid_qops_start_streaming,
>>>>> +	.stop_streaming = isp4vid_qops_stop_streaming,
>>>>> +	.wait_prepare = vb2_ops_wait_prepare,
>>>>> +	.wait_finish = vb2_ops_wait_finish,
>>>>> +};
>>>>
>>>> [snip]
>>>>
>>>> Along with the changes from my other emails, I believe this finally fixes all of
>>>> the refcounting issues. No more UaF or leaks here. :-)
>>>>
>>>> Sultan
>>> Tried changes in this and other two mails together, found the kernel will
>>> crash when re-open cheese App (means open cheese, close it and then open it
>>> again), the crash stack dump is shown below
>>>
>>> [   38.008686] amdgpu 0000:c3:00.0: amdgpu: power on isp suc
>>> [   38.032693] amdgpu 0000:c3:00.0: amdgpu: set xclk 788 suc, reg_val 800
>>> [   38.060866] amdgpu 0000:c3:00.0: amdgpu: set iclk 788 suc, reg_val 792
>>> [   38.060877] amdgpu 0000:c3:00.0: amdgpu: set xclk 788 iclk 788 to
>>> performance state 1 suc
>>> [   38.425284] amdgpu 0000:c3:00.0: amdgpu: isp set performance do none to
>>> state 0
>>> [   38.448631] amdgpu 0000:c3:00.0: amdgpu: power off isp suc
>>> [   38.508368] amdgpu 0000:c3:00.0: amdgpu: power on isp suc
>>> [   38.538525] amdgpu 0000:c3:00.0: amdgpu: set xclk 788 suc, reg_val 800
>>> [   38.568839] amdgpu 0000:c3:00.0: amdgpu: set iclk 788 suc, reg_val 792
>>> [   38.568851] amdgpu 0000:c3:00.0: amdgpu: set xclk 788 iclk 788 to
>>> performance state 1 suc
>>> [   38.997779] amdgpu 0000:c3:00.0: amdgpu: isp set performance do none to
>>> state 0
>>> [   39.029297] amdgpu 0000:c3:00.0: amdgpu: power off isp suc
>>> [   39.084010] BUG: kernel NULL pointer dereference, address:
>>> 0000000000000008
>>> [   39.084024] #PF: supervisor read access in kernel mode
>>> [   39.084028] #PF: error_code(0x0000) - not-present page
>>> [   39.084031] PGD 0 P4D 0
>>> [   39.084037] Oops: Oops: 0000 [#1] PREEMPT SMP NOPTI
>>> [   39.084042] CPU: 3 UID: 1000 PID: 3119 Comm: viewfinderbin-q Not tainted
>>> 6.14.0-rc2+ #944
>>> [   39.084046] Hardware name: HP HP ZBook Ultra G1a 14 inch Mobile
>>> Workstation PC/8D01, BIOS X89 Ver. 01.03.00 04/25/2025
>>> [   39.084049] RIP: 0010:validate_page_before_insert+0x5/0xc0
>>> [   39.084060] Code: 31 c9 31 ff 45 31 c0 c3 cc cc cc cc 66 0f 1f 84 00 00
>>> 00 00 00 90 90 90 90 90 90 90 90 90 90 90 90 90 90 90 90 0f 1f 44 00 00 <48>
>>> 8b 46 08 a8 01 75 7a 66 90 48 89 f0 8b 50 34 85 d2 0f 84 8c 00
>>> [   39.084064] RSP: 0018:ffffb95e05e7b6a0 EFLAGS: 00010202
>>> [   39.084067] RAX: 0000000000000000 RBX: 8000000000000025 RCX:
>>> 0000000000000000
>>> [   39.084069] RDX: 0000000000000000 RSI: 0000000000000000 RDI:
>>> ffff9efa10e400b8
>>> [   39.084071] RBP: ffffb95e05e7b6e0 R08: 0000000000000000 R09:
>>> 0000000000000000
>>> [   39.084072] R10: 0000000000000000 R11: 0000000000000000 R12:
>>> ffff9efa10e400b8
>>> [   39.084074] R13: ffffb95e0e055000 R14: 00007fa3d6119000 R15:
>>> 0000000000000000
>>> [   39.084076] FS:  00007fa3f4bf96c0(0000) GS:ffff9f050d780000(0000)
>>> knlGS:0000000000000000
>>> [   39.084078] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
>>> [   39.084080] CR2: 0000000000000008 CR3: 0000000137bba000 CR4:
>>> 0000000000750ef0
>>> [   39.084083] PKRU: 55555554
>>> [   39.084084] Call Trace:
>>> [   39.084087]  <TASK>
>>> [   39.084093]  ? show_regs+0x71/0x90
>>> [   39.084099]  ? __die+0x25/0x80
>>> [   39.084102]  ? page_fault_oops+0x15c/0x570
>>> [   39.084108]  ? do_user_addr_fault+0x4a9/0x810
>>> [   39.084111]  ? __mod_memcg_lruvec_state+0xde/0x280
>>> [   39.084117]  ? exc_page_fault+0x83/0x1d0
>>> [   39.084123]  ? asm_exc_page_fault+0x27/0x30
>>> [   39.084131]  ? validate_page_before_insert+0x5/0xc0
>>> [   39.084134]  ? vm_insert_page+0xd4/0x1a0
>>> [   39.084139]  remap_vmalloc_range_partial+0xe3/0x160
>>> [   39.084145]  remap_vmalloc_range+0x21/0x40
>>> [   39.084153]  isp4vid_vb2_mmap+0x2d/0x130 [amd_capture]
>>> [   39.084164]  isp4vid_vb2_dmabuf_ops_mmap+0x12/0x20 [amd_capture]
>>> [   39.084168]  dma_buf_mmap_internal+0x41/0x70
>>> [   39.084173]  __mmap_region+0x62f/0xbc0
>>> [   39.084179]  mmap_region+0x67/0xe0
>>> [   39.084181]  do_mmap+0x52b/0x650
>>> [   39.084187]  vm_mmap_pgoff+0xf4/0x1c0
>>> [   39.084194]  ksys_mmap_pgoff+0x182/0x250
>>> [   39.084198]  __x64_sys_mmap+0x33/0x70
>>> [   39.084202]  x64_sys_call+0x2541/0x26f0
>>> [   39.084206]  do_syscall_64+0x70/0x130
>>> [   39.084212]  ? __rseq_handle_notify_resume+0xa4/0x520
>>> [   39.084216]  ? do_futex+0x17f/0x220
>>> [   39.084220]  ? restore_fpregs_from_fpstate+0x3d/0xd0
>>> [   39.084225]  ? switch_fpu_return+0x50/0xe0
>>> [   39.084229]  ? syscall_exit_to_user_mode+0x18c/0x1c0
>>> [   39.084234]  ? do_syscall_64+0x7c/0x130
>>> [   39.084237]  ? __pfx_futex_wake_mark+0x10/0x10
>>> [   39.084241]  ? hrtimer_cancel+0x15/0x50
>>> [   39.084246]  ? futex_wait+0x7c/0x120
>>> [   39.084249]  ? __pfx_hrtimer_wakeup+0x10/0x10
>>> [   39.084254]  ? __rseq_handle_notify_resume+0xa4/0x520
>>> [   39.084256]  ? do_futex+0x17f/0x220
>>> [   39.084258]  ? restore_fpregs_from_fpstate+0x3d/0xd0
>>> [   39.084261]  ? switch_fpu_return+0x50/0xe0
>>> [   39.084264]  ? syscall_exit_to_user_mode+0x18c/0x1c0
>>> [   39.084268]  ? do_syscall_64+0x7c/0x130
>>> [   39.084272]  ? __count_memcg_events+0xcd/0x190
>>> [   39.084276]  ? count_memcg_events.constprop.0+0x2a/0x50
>>> [   39.084279]  ? handle_mm_fault+0x1ef/0x2d0
>>> [   39.084284]  ? do_user_addr_fault+0x58d/0x810
>>> [   39.084286]  ? irqentry_exit_to_user_mode+0x33/0x180
>>> [   39.084291]  ? irqentry_exit+0x43/0x50
>>> [   39.084294]  ? exc_page_fault+0x94/0x1d0
>>> [   39.084298]  entry_SYSCALL_64_after_hwframe+0x76/0x7e
>>> [   39.084302] RIP: 0033:0x7fa463b2531c
>>> [   39.084305] Code: 1e fa 41 f7 c1 ff 0f 00 00 75 33 55 48 89 e5 41 54 41
>>> 89 cc 53 48 89 fb 48 85 ff 74 41 45 89 e2 48 89 df b8 09 00 00 00 0f 05 <48>
>>> 3d 00 f0 ff ff 77 7c 5b 41 5c 5d c3 0f 1f 80 00 00 00 00 48 8b
>>> [   39.084308] RSP: 002b:00007fa3f4bf79e0 EFLAGS: 00000246 ORIG_RAX:
>>> 0000000000000009
>>> [   39.084311] RAX: ffffffffffffffda RBX: 0000000000000000 RCX:
>>> 00007fa463b2531c
>>> [   39.084312] RDX: 0000000000000001 RSI: 00000000007782c0 RDI:
>>> 0000000000000000
>>> [   39.084314] RBP: 00007fa3f4bf79f0 R08: 0000000000000047 R09:
>>> 0000000000000000
>>> [   39.084315] R10: 0000000000000001 R11: 0000000000000246 R12:
>>> 0000000000000001
>>> [   39.084317] R13: 0000000000000000 R14: 00007fa3f8019b48 R15:
>>> 00007fa3f4bf7b60
>>> [   39.084320]  </TASK>
>>> [   39.084321] Modules linked in: snd_seq_dummy snd_hrtimer hid_sensor_prox
>>> hid_sensor_gyro_3d hid_sensor_trigger industrialio_triggered_buffer
>>> kfifo_buf hid_sensor_iio_common industrialio hid_sensor_hub rfcomm cmac
>>> algif_hash algif_skcipher af_alg hid_logitech_hidpp ov05c10 v4l2_cci
>>> amd_capture v4l2_fwnode videobuf2_memops videobuf2_v4l2 videobuf2_common
>>> v4l2_async videodev pinctrl_amdisp i2c_designware_amdisp qrtr bnep btusb
>>> btrtl btintel btbcm btmtk bluetooth hid_logitech_dj intel_rapl_msr amd_atl
>>> intel_rapl_common binfmt_misc nls_iso8859_1 snd_acp_legacy_mach snd_acp_mach
>>> snd_soc_nau8821 snd_acp3x_rn amdgpu edac_mce_amd snd_ctl_led
>>> snd_hda_codec_realtek snd_acp70 snd_hda_codec_generic snd_acp_i2s
>>> snd_acp_pdm snd_hda_scodec_component snd_soc_dmic snd_acp_pcm
>>> snd_sof_amd_rembrandt snd_sof_amd_renoir snd_sof_amd_acp snd_sof_pci
>>> snd_sof_xtensa_dsp snd_hda_codec_hdmi snd_sof snd_sof_utils snd_pci_ps
>>> snd_soc_acpi_amd_match snd_amd_sdw_acpi soundwire_amd
>>> soundwire_generic_allocation soundwire_bus snd_hda_intel kvm_amd
>>> [   39.084385]  snd_soc_sdca snd_intel_dspcfg snd_intel_sdw_acpi
>>> snd_soc_core snd_hda_codec amdxcp kvm snd_compress
>>> drm_panel_backlight_quirks snd_hda_core snd_usb_audio gpu_sched ac97_bus
>>> polyval_clmulni snd_pcm_dmaengine polyval_generic ghash_clmulni_intel
>>> drm_buddy snd_usbmidi_lib snd_seq_midi snd_rpl_pci_acp6x sha256_ssse3
>>> snd_ump drm_ttm_helper snd_seq_midi_event snd_hwdep sha1_ssse3 ttm
>>> aesni_intel drm_exec snd_rawmidi drm_suballoc_helper snd_acp_pci mc
>>> crypto_simd snd_acp_legacy_common drm_client_lib cryptd drm_display_helper
>>> snd_pci_acp6x snd_seq rapl hp_wmi cec wmi_bmof snd_pcm sparse_keymap
>>> snd_seq_device rc_core snd_timer snd_pci_acp5x drm_kms_helper
>>> snd_rn_pci_acp3x amd_pmf snd i2c_algo_bit snd_acp_config i2c_piix4
>>> snd_soc_acpi amdtee soundcore snd_pci_acp3x i2c_smbus ccp joydev tee
>>> input_leds serial_multi_instantiate amd_pmc platform_profile wireless_hotkey
>>> mac_hid serio_raw sch_fq_codel msr parport_pc ppdev lp parport efi_pstore
>>> nfnetlink dmi_sysfs ip_tables x_tables autofs4 r8152 mii usbhid
>>> hid_multitouch
>>> [   39.084451]  hid_generic nvme ucsi_acpi i2c_hid_acpi video typec_ucsi
>>> i2c_hid amd_sfh thunderbolt hid nvme_core typec wmi drm
>>> [   39.084465] CR2: 0000000000000008
>>> [   39.084468] ---[ end trace 0000000000000000 ]---
>>> [   39.116114] amdgpu 0000:c3:00.0: amdgpu: power on isp suc
>>> [   39.159519] amdgpu 0000:c3:00.0: amdgpu: set xclk 788 suc, reg_val 800
>>> [   39.201325] amdgpu 0000:c3:00.0: amdgpu: set iclk 788 suc, reg_val 792
>>> [   39.201334] amdgpu 0000:c3:00.0: amdgpu: set xclk 788 iclk 788 to
>>> performance state 1 suc
>>> [   39.343894] RIP: 0010:validate_page_before_insert+0x5/0xc0
>>> [   39.343907] Code: 31 c9 31 ff 45 31 c0 c3 cc cc cc cc 66 0f 1f 84 00 00
>>> 00 00 00 90 90 90 90 90 90 90 90 90 90 90 90 90 90 90 90 0f 1f 44 00 00 <48>
>>> 8b 46 08 a8 01 75 7a 66 90 48 89 f0 8b 50 34 85 d2 0f 84 8c 00
>>> [   39.343909] RSP: 0018:ffffb95e05e7b6a0 EFLAGS: 00010202
>>> [   39.343911] RAX: 0000000000000000 RBX: 8000000000000025 RCX:
>>> 0000000000000000
>>> [   39.343913] RDX: 0000000000000000 RSI: 0000000000000000 RDI:
>>> ffff9efa10e400b8
>>> [   39.343913] RBP: ffffb95e05e7b6e0 R08: 0000000000000000 R09:
>>> 0000000000000000
>>> [   39.343914] R10: 0000000000000000 R11: 0000000000000000 R12:
>>> ffff9efa10e400b8
>>> [   39.343915] R13: ffffb95e0e055000 R14: 00007fa3d6119000 R15:
>>> 0000000000000000
>>> [   39.343916] FS:  00007fa3f4bf96c0(0000) GS:ffff9f050d780000(0000)
>>> knlGS:0000000000000000
>>> [   39.343917] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
>>> [   39.343918] CR2: 0000000000000008 CR3: 0000000137bba000 CR4:
>>> 0000000000750ef0
>>> [   39.343919] PKRU: 55555554
>>> [   39.343920] note: viewfinderbin-q[3119] exited with irqs disabled
>>>
>>> This is the patch, it should have contained all your changes
>>>
>>>  From b0a61b6b7500635928166516d5563e6de50300bb Mon Sep 17 00:00:00 2001
>>> From: Bin Du <Bin.Du@....com>
>>> Date: Mon, 28 Jul 2025 14:18:04 +0800
>>> Subject: [PATCH] add comment from community
>>>
>>> Change-Id: I06fff629c0dc691eaeaa9d9b81e6af27cbb6adb2
>>> ---
>>>   drivers/media/platform/amd/isp4/isp4_video.c | 59 ++++++--------------
>>>   1 file changed, 17 insertions(+), 42 deletions(-)
>>>
>>> diff --git a/drivers/media/platform/amd/isp4/isp4_video.c
>>> b/drivers/media/platform/amd/isp4/isp4_video.c
>>> index a1376498c246..b210a0b1e49b 100644
>>> --- a/drivers/media/platform/amd/isp4/isp4_video.c
>>> +++ b/drivers/media/platform/amd/isp4/isp4_video.c
>>> @@ -183,7 +183,7 @@ static int isp4vid_vb2_mmap(void *buf_priv, struct
>>> vm_area_struct *vma)
>>>   	vm_flags_set(vma, VM_DONTEXPAND);
>>>
>>>   	dev_dbg(buf->dev, "mmap isp user bo 0x%llx size %ld refcount %d\n",
>>> -		buf->gpu_addr, buf->size, buf->refcount.refs.counter);
>>> +		buf->gpu_addr, buf->size, refcount_read(&buf->refcount));
>>>
>>>   	return 0;
>>>   }
>>> @@ -214,14 +214,9 @@ static void isp4vid_vb2_detach_dmabuf(void *mem_priv)
>>>   		return;
>>>   	}
>>>
>>> -	struct iosys_map map = IOSYS_MAP_INIT_VADDR(buf->vaddr);
>>> -
>>>   	dev_dbg(buf->dev, "detach dmabuf of isp user bo 0x%llx size %ld",
>>>   		buf->gpu_addr, buf->size);
>>>
>>> -	if (buf->vaddr)
>>> -		dma_buf_vunmap_unlocked(buf->dbuf, &map);
>>> -
>>>   	kfree(buf);
>>>   }
>>>
>>> @@ -470,21 +465,25 @@ static struct dma_buf *isp4vid_vb2_get_dmabuf(struct
>>> vb2_buffer *vb,
>>>   					      unsigned long flags)
>>>   {
>>>   	struct isp4vid_vb2_buf *buf = buf_priv;
>>> -	struct dma_buf *dbuf;
>>> +	struct dma_buf *dbuf = buf->dbuf;
>>> +
>>> +	if (dbuf) {
>>> +		dev_dbg(buf->dev, "dbuf already created, reuse %s dbuf\n",
>>> +			buf->is_expbuf ? "exported" : "implicit");
>>> +		get_dma_buf(dbuf);
>>>
>>> -	if (buf->dbuf) {
>>> -		dev_dbg(buf->dev, "dbuf already created, reuse implicit dbuf\n");
>>> -		dbuf = buf->dbuf;
>>>   	} else {
>>>   		dbuf = isp4vid_get_dmabuf(vb, buf_priv, flags);
>>> +		if (!dbuf)
>>> +			return NULL;
>>> +
>>>   		dev_dbg(buf->dev, "created new dbuf\n");
>>> +		buf->is_expbuf = true;
>>> +		refcount_inc(&buf->refcount);
>>>   	}
>>>
>>> -	buf->is_expbuf = true;
>>> -	refcount_inc(&buf->refcount);
>>> -
>>>   	dev_dbg(buf->dev, "buf exported, refcount %d\n",
>>> -		buf->refcount.refs.counter);
>>> +		refcount_read(&buf->refcount));
>>>
>>>   	return dbuf;
>>>   }
>>> @@ -579,8 +578,9 @@ static void isp4vid_vb2_put(void *buf_priv)
>>>   {
>>>   	struct isp4vid_vb2_buf *buf = (struct isp4vid_vb2_buf *)buf_priv;
>>>   	unsigned int ref_cnt = refcount_read(&buf->refcount);
>>> +	struct device *dev = buf->dev;
>>>
>>> -	dev_dbg(buf->dev,
>>> +	dev_dbg(dev,
>>>   		"release isp user bo 0x%llx size %ld refcount %u is_expbuf %d\n",
>>>   		buf->gpu_addr, buf->size,
>>>   		ref_cnt, buf->is_expbuf);
>>> @@ -598,7 +598,7 @@ static void isp4vid_vb2_put(void *buf_priv)
>>>   		/* the ref_cnt - 1 is just a predicted value as reference, can't
>>>   		 * guarantee it's the actual value now
>>>   		 */
>>> -		dev_warn(buf->dev, "ignore buffer free, refcount %u > 0\n", ref_cnt - 1);
>>> +		dev_warn(dev, "ignore buffer free, refcount %u > 0\n", ref_cnt - 1);
>>>   	}
>>>   }
>>>
>>> @@ -642,7 +642,7 @@ static void *isp4vid_vb2_alloc(struct vb2_buffer *vb,
>>> struct device *dev,
>>>   	refcount_set(&buf->refcount, 1);
>>>
>>>   	dev_dbg(dev, "allocated isp user bo 0x%llx size %ld refcount %d\n",
>>> -		buf->gpu_addr, buf->size, buf->refcount.refs.counter);
>>> +		buf->gpu_addr, buf->size, refcount_read(&buf->refcount));
>>>
>>>   	return buf;
>>>   err_free:
>>> @@ -1140,30 +1140,6 @@ static void isp4vid_qops_buffer_queue(struct
>>> vb2_buffer *vb)
>>>   	mutex_unlock(&isp_vdev->buf_list_lock);
>>>   }
>>>
>>> -static void isp4vid_qops_buffer_cleanup(struct vb2_buffer *vb)
>>> -{
>>> -	struct isp4vid_dev *isp_vdev = vb2_get_drv_priv(vb->vb2_queue);
>>> -	struct isp4vid_vb2_buf *buf = vb->planes[0].mem_priv;
>>> -
>>> -	dev_dbg(isp_vdev->dev, "%s|index=%u vb->memory %u",
>>> -		isp_vdev->vdev.name, vb->index, vb->memory);
>>> -
>>> -	if (!buf) {
>>> -		dev_err(isp_vdev->dev, "Invalid buf handle");
>>> -		return;
>>> -	}
>>> -
>>> -	// release implicit dmabuf reference here for vb2 buffer
>>> -	// of type MMAP and is exported
>>> -	if (vb->memory == VB2_MEMORY_MMAP && buf->is_expbuf) {
>>> -		dma_buf_put(buf->dbuf);
>>> -		dev_dbg(isp_vdev->dev,
>>> -			"put dmabuf for vb->memory %d expbuf %d",
>>> -			vb->memory,
>>> -			buf->is_expbuf);
>>> -	}
>>> -}
>>> -
>>>   static int isp4vid_qops_start_streaming(struct vb2_queue *vq,
>>>   					unsigned int count)
>>>   {
>>> @@ -1343,7 +1319,6 @@ static int isp4vid_fill_buffer_size(struct isp4vid_dev
>>> *isp_vdev)
>>>
>>>   static const struct vb2_ops isp4vid_qops = {
>>>   	.queue_setup = isp4vid_qops_queue_setup,
>>> -	.buf_cleanup = isp4vid_qops_buffer_cleanup,
>>>   	.buf_queue = isp4vid_qops_buffer_queue,
>>>   	.start_streaming = isp4vid_qops_start_streaming,
>>>   	.stop_streaming = isp4vid_qops_stop_streaming,
>>> -- 
>>> 2.34.1
>>
>> The patch looks correct. I will get back to you on this after doing some testing
>> myself.
>>
>> FYI, I can only test these changes with the old isp4 driver right now, since the
>> new isp4 driver doesn't work for me as mentioned in my other email. So far,
>> something does seem wrong after these changes I suggested because they break the
>> camera on the old isp4 driver too.
> 
> Following up on this: there were several more issues hiding in the driver that
> needed to be fixed unfortunately. I ended up making a lot of changes to fix all
> of the buffer lifetime issues, and I created individual commits for this work
> because it was getting hard to keep track of all the changes.
> 
> I think I'll have this completed tomorrow and will provide a link to my tree
> with my commits. Hopefully your internal branch hasn't diverged too much from
> this patchset submission, otherwise there may be some conflicts to resolve.
> 
> Sultan

Really appreciate your time and effort. Yes, big code changes happen in 
our internal branch to address upstream comments, but i believe your 
commits will be extremely valuable. We can check and test after fixing 
possbile conflict.

Regards,
Bin

Powered by blists - more mailing lists