[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <alpine.LRH.2.00.1610041519160.1123@gjva.wvxbf.pm>
Date: Tue, 4 Oct 2016 15:26:28 +0200 (CEST)
From: Jiri Kosina <jikos@...nel.org>
To: Jörg Otte <jrg.otte@...il.com>
cc: Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
Andy Lutomirski <luto@...nel.org>,
Linus Torvalds <torvalds@...ux-foundation.org>,
Ingo Molnar <mingo@...nel.org>,
Michael Krufky <mkrufky@...uxtv.org>,
Mauro Carvalho Chehab <mchehab@...nel.org>,
Patrick Boettcher <patrick.boettcher@...teo.de>,
linux-media@...r.kernel.org
Subject: Re: Problem with VMAP_STACK=y
On Tue, 4 Oct 2016, Jörg Otte wrote:
> With kernel 4.8.0-01558-g21f54dd I get thousands of
> "dvb-usb: bulk message failed: -11 (1/0)"
> messages in the logs and the DVB adapter is not working.
>
> It tourned out the new config option VMAP_STACK=y (which is the default)
> is the culprit.
> No problems for me with VMAP_STACK=n.
I'd guess that this is EAGAIN coming from usb_hcd_map_urb_for_dma() as the
DVB driver is trying to perform on-stack DMA.
Not really knowing which driver exactly you're using, I quickly skimmed
through DVB sources, and it turns out this indeed seems to be rather
common antipattern, and it should be fixed nevertheless. See
cxusb_ctrl_msg()
dibusb_power_ctrl()
dibusb2_0_streaming_ctrl()
dibusb2_0_power_ctrl()
digitv_ctrl_msg()
dtt200u_fe_init()
dtt200u_fe_set_frontend()
dtt200u_power_ctrl()
dtt200u_streaming_ctrl()
dtt200u_pid_filter()
Adding relevant CCs.
--
Jiri Kosina
SUSE Labs
Powered by blists - more mailing lists