linux-kernel - Re: [PATCH] ALSA: core: Fix control device release issue

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives
Hash Suite for Android: free password hash cracker in your pocket
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <s5ho8q1hbtl.wl-tiwai@suse.de>
Date:   Tue, 02 Jun 2020 14:49:42 +0200
From:   Takashi Iwai <tiwai@...e.de>
To:     朱灿灿 <zhucancan@...o.com>
Cc:     <tiwai@...e.com>, <perex@...ex.cz>, <alsa-devel@...a-project.org>,
        <linux-kernel@...r.kernel.org>, kernel <kernel@...o.com>,
        "wenhu wang" <wenhu.wang@...o.com>
Subject: Re: [PATCH] ALSA: core: Fix control device release issue

On Tue, 02 Jun 2020 11:48:04 +0200,
朱灿灿 wrote:
> 
> From: zhucancan <zhucancan@...o.com>
> Date: Tue, 2 Jun 2020 16:22:57 +0800
> Subject: [PATCH] ALSA: core: Fix control device release issue
> 
> We use snd_pcm_add_usr_ctls() in component's .pcm_new(),
> unfortunately snd_soc_dapm_add_routes() meets error during
> adding card->dapm_routes/of_dapm_routes, it will goto probe_end
> to call soc_cleanup_card_resources().
> 
> The commit dc82e52492f6 ("ALSA: core: Assure control device
> to be registered at last") will make pcm device release is
> prior to control device release, but the control device needs
> to use pcm pointor, which is already freed by pcm device release.
> 
> [   70.056000] Unable to handle kernel paging request at virtual address ffffff80f2d2c480
> [   70.066139] init: processing action (init.svc.media=*) from (/system/etc/init/mediaserver.rc:1)
> [   70.076311] Mem abort info:
> [   70.126679]   ESR = 0x96000047
> [   70.126685]   EC = 0x25: DABT (current EL), IL = 32 bits
> [   70.126689]   SET = 0, FnV = 0
> [   70.126693]   EA = 0, S1PTW = 0
> [   70.126696] Data abort info:
> [   70.126700]   ISV = 0, ISS = 0x00000047
> [   70.126704]   CM = 0, WnR = 1
> [   70.126710] swapper pgtable: 4k pages, 39-bit VAs, pgdp=00000000a216b000
> [   70.126715] [ffffff80f2d2c480] pgd=00000001ffb7f003, pud=00000001ffb7f003, pmd=00000001ff9e8003, pte=0068000172d2cf12
> [   70.126731] Internal error: Oops: 96000047 [#1] PREEMPT SMP
> [   70.126821] CPU: 0 PID: 5 Comm: kworker/0:0 Tainted: G        W  O      5.4.12-qgki-debug-g4c7ca9f20 #10
> [   70.126840] Workqueue: events deferred_probe_work_func$2b500e852cead69fde05a2b98ceb2fa6.cfi_jt
> [   70.126845] pstate: 80c00005 (Nzcv daif +PAN +UAO)
> [   70.126854] pc : pcm_usr_ctl_private_free$165f4400ee6732c7c8c8157be0f72518+0x20/0x34
> [   70.126861] lr : snd_ctl_remove+0xf4/0x154
> [   70.126862] sp : ffffffc01007b750
> [   70.126865] x29: ffffffc01007b750 x28: ffffffd973c2bb2c
> [   70.126868] x27: ffffff8107b741a0 x26: dead000000000100
> [   70.126874] x25: 0000000000000002 x24: ffffffd973c2bb8c
> [   70.126877] x23: ffffffd975027c80 x22: ffffffd97415437f
> [   70.126880] x21: 0000000000000001 x20: ffffffd973c37e10
> [   70.126883] x19: ffffff80f2d55d00 x18: ffffff80451de4d0
> [   70.126886] x17: ffffff80f2d58000 x16: 0000000000000400
> [   70.126889] x15: a6fe4e9bdc5a6b05 x14: 0000000000000000
> [   70.126892] x13: ffffffff03ab5408 x12: 0000000000000001
> [   70.126896] x11: 7079542070704120 x10: 0000000000000398
> [   70.126899] x9 : 0000000000000001 x8 : ffffff80f2d2c398
> [   70.126902] x7 : 0000000000000000 x6 : ffffff80f6b4d178
> [   70.126905] x5 : ffffffd97335fe8c x4 : 0000000000000000
> [   70.126908] x3 : ffffffc01007b650 x2 : ffffffc01007b768
> [   70.126911] x1 : 00000000ffffffff x0 : ffffff811ff41480
> [   70.126916] Call trace:
> [   70.126920]  pcm_usr_ctl_private_free$165f4400ee6732c7c8c8157be0f72518+0x20/0x34

Which driver or core contains this function?  I can't find it in the
latest upstream code.

Overall, it looks rather like an issue of the driver side, not the
ALSA core.  If the relevant control is tied with a PCM stream, it has
to be freed individually beforehand at the time when the stream gets
freed.


thanks,

Takashi

> [   70.126923]  snd_ctl_remove+0xf4/0x154
> [   70.126927]  snd_ctl_dev_free$1b8efea49186b79e6e4a39cac5748f1c+0x3c/0x70
> [   70.126930]  snd_device_free_all+0x234/0x2c8
> [   70.126934]  release_card_device$fb7328386cac0bdf1b3f2f88da79521f+0x24/0xcc
> [   70.126941]  device_release$d2fdd8912beb71f9ccc7ca72fceb0522+0x48/0x118
> [   70.126949]  kobject_cleanup+0x138/0x25c
> [   70.126952]  kobject_put+0x50/0x60
> [   70.126955]  put_device+0x14/0x20
> [   70.126957]  snd_card_free+0x60/0x98
> [   70.126963]  soc_cleanup_card_resources+0x2c/0x420
> [   70.126966]  snd_soc_bind_card+0xcc0/0xfa8
> [   70.126968]  snd_soc_register_card+0x110/0x128
> [   70.126974]  devm_snd_soc_register_card+0x4c/0x90
> 
> Fixes: dc82e52492f6 ("ALSA: core: Assure control device to be registered at last")
> Signed-off-by: zhucancan <zhucancan@...o.com>
> ---
>  sound/core/device.c | 8 --------
>  1 file changed, 8 deletions(-)
> 
> diff --git a/sound/core/device.c b/sound/core/device.c
> index bf0b04a7ee79..1aee0f5623a2 100644
> --- a/sound/core/device.c
> +++ b/sound/core/device.c
> @@ -225,14 +225,6 @@ void snd_device_free_all(struct snd_card *card)
>  
>  	if (snd_BUG_ON(!card))
>  		return;
> -	list_for_each_entry_safe_reverse(dev, next, &card->devices, list) {
> -		/* exception: free ctl and lowlevel stuff later */
> -		if (dev->type == SNDRV_DEV_CONTROL ||
> -		    dev->type == SNDRV_DEV_LOWLEVEL)
> -			continue;
> -		__snd_device_free(dev);
> -	}
> -
>  	/* free all */
>  	list_for_each_entry_safe_reverse(dev, next, &card->devices, list)
>  		__snd_device_free(dev);
> -- 
> 2.21.0
> 
> 
> 
> 
> 
> 
> 
> 
>