From: Alan Stern <stern@rowland.harvard.edu> To: Oliver Neukum <oliver@neukum.org> Cc: Daniel Mack <daniel@caiaq.de>, <linux-kernel@vger.kernel.org>, Pedro Ribeiro <pedrib@gmail.com>, <akpm@linux-foundation.org>, Greg KH <gregkh@suse.de>, <alsa-devel@alsa-project.org>, <linux-usb@vger.kernel.org> Subject: Re: USB transfer_buffer allocations on 64bit systems Date: Fri, 9 Apr 2010 10:41:48 -0400 (EDT) [thread overview] Message-ID: <Pine.LNX.4.44L0.1004091033150.1852-100000@iolanthe.rowland.org> (raw) In-Reply-To: <201004090804.36213.oliver@neukum.org> On Fri, 9 Apr 2010, Oliver Neukum wrote: > Am Freitag, 9. April 2010 00:20:36 schrieb Alan Stern: > > > > That would work, but it doesn't match the way existing drivers use the > > > > interface. For example, the audio driver allocates a 16-byte coherent > > > > buffer and then uses four bytes from it for each of four different > > > > URBs. > > > > > > That will not work with any fallback that does not yield a coherent buffer. > > > > What you mean isn't entirely clear. But it certainly does work in > > various circumstances that don't yield coherent buffers. For example, > > it works if the controller uses PIO instead of DMA. It also works if > > the controller uses DMA and the URBs have to be bounced. > > It'll work on x86. On incoherent architectures this violates the cacheline > rules for DMA-mapping if you have to bounce. Not true. Consider: The driver allocates a 16-byte buffer (xbuf) divided up into four sets of four bytes, and sets urb[i].transfer_buffer_dma = xbuf_dma + 4*i; Then usb_submit_urb(urb[i]) will copy the appropriate four bytes to a bounce buffer and map the bounce buffer. Accesses to the other parts of xbuf won't violate the cacheline rules, because xbuf isn't mapped for DMA -- only the bounce buffer is. When urb[i] completes, the bounce buffer contents will be copied back to the original four bytes in xbuf. Again, there is no violation of cacheline rules. > So it seems to me that > if you want to share a buffer between URBs, it must be coherent. No. But it must be allocated via usb_alloc_buffer() (or whatever that routine gets renamed to). Alan Stern
WARNING: multiple messages have this Message-ID (diff)
From: Alan Stern <stern-nwvwT67g6+6dFdvTe/nMLpVzexx5G7lz@public.gmane.org> To: Oliver Neukum <oliver-GvhC2dPhHPQdnm+yROfE0A@public.gmane.org> Cc: Daniel Mack <daniel-rDUAYElUppE@public.gmane.org>, linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, Pedro Ribeiro <pedrib-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>, akpm-de/tnXTf+JLsfHDXvbKv3WD2FQJk+8+b@public.gmane.org, Greg KH <gregkh-l3A5Bk7waGM@public.gmane.org>, alsa-devel-K7yf7f+aM1XWsZ/bQMPhNw@public.gmane.org, linux-usb-u79uwXL29TY76Z2rM5mHXA@public.gmane.org Subject: Re: USB transfer_buffer allocations on 64bit systems Date: Fri, 9 Apr 2010 10:41:48 -0400 (EDT) [thread overview] Message-ID: <Pine.LNX.4.44L0.1004091033150.1852-100000@iolanthe.rowland.org> (raw) In-Reply-To: <201004090804.36213.oliver-GvhC2dPhHPQdnm+yROfE0A@public.gmane.org> On Fri, 9 Apr 2010, Oliver Neukum wrote: > Am Freitag, 9. April 2010 00:20:36 schrieb Alan Stern: > > > > That would work, but it doesn't match the way existing drivers use the > > > > interface. For example, the audio driver allocates a 16-byte coherent > > > > buffer and then uses four bytes from it for each of four different > > > > URBs. > > > > > > That will not work with any fallback that does not yield a coherent buffer. > > > > What you mean isn't entirely clear. But it certainly does work in > > various circumstances that don't yield coherent buffers. For example, > > it works if the controller uses PIO instead of DMA. It also works if > > the controller uses DMA and the URBs have to be bounced. > > It'll work on x86. On incoherent architectures this violates the cacheline > rules for DMA-mapping if you have to bounce. Not true. Consider: The driver allocates a 16-byte buffer (xbuf) divided up into four sets of four bytes, and sets urb[i].transfer_buffer_dma = xbuf_dma + 4*i; Then usb_submit_urb(urb[i]) will copy the appropriate four bytes to a bounce buffer and map the bounce buffer. Accesses to the other parts of xbuf won't violate the cacheline rules, because xbuf isn't mapped for DMA -- only the bounce buffer is. When urb[i] completes, the bounce buffer contents will be copied back to the original four bytes in xbuf. Again, there is no violation of cacheline rules. > So it seems to me that > if you want to share a buffer between URBs, it must be coherent. No. But it must be allocated via usb_alloc_buffer() (or whatever that routine gets renamed to). Alan Stern -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org More majordomo info at http://vger.kernel.org/majordomo-info.html
next prev parent reply other threads:[~2010-04-09 14:42 UTC|newest] Thread overview: 221+ messages / expand[flat|nested] mbox.gz Atom feed top 2010-04-07 9:06 USB transfer_buffer allocations on 64bit systems Daniel Mack 2010-04-07 9:06 ` Daniel Mack [not found] ` <p2g581ef6d61004070220z1153d40ez955b356e01220848@mail.gmail.com> 2010-04-07 9:26 ` USB HID gadget driver (was: Re: USB transfer_buffer allocations on 64bit systems) Daniel Mack 2010-04-07 9:26 ` Daniel Mack 2010-04-07 14:59 ` USB transfer_buffer allocations on 64bit systems Alan Stern 2010-04-07 14:59 ` Alan Stern 2010-04-07 15:11 ` Daniel Mack 2010-04-07 15:11 ` Daniel Mack 2010-04-07 15:31 ` Greg KH 2010-04-07 15:31 ` Greg KH 2010-04-07 15:35 ` Daniel Mack 2010-04-07 15:35 ` Daniel Mack 2010-04-07 15:51 ` Greg KH 2010-04-07 15:51 ` Greg KH 2010-04-07 16:04 ` Alan Stern 2010-04-07 16:04 ` Alan Stern 2010-04-08 6:09 ` Oliver Neukum 2010-04-08 11:07 ` Daniel Mack 2010-04-08 11:07 ` Daniel Mack 2010-04-07 15:55 ` Alan Stern 2010-04-07 15:55 ` Alan Stern 2010-04-07 16:16 ` Daniel Mack 2010-04-07 16:16 ` Daniel Mack 2010-04-07 16:47 ` Alan Stern 2010-04-07 16:47 ` Alan Stern 2010-04-07 17:55 ` Takashi Iwai 2010-04-07 17:55 ` Takashi Iwai 2010-04-07 17:59 ` Daniel Mack 2010-04-07 17:59 ` Daniel Mack 2010-04-07 18:06 ` Takashi Iwai 2010-04-07 18:06 ` Takashi Iwai 2010-04-07 19:13 ` Alan Stern 2010-04-07 19:13 ` Alan Stern 2010-04-07 23:59 ` Robert Hancock 2010-04-07 23:59 ` Robert Hancock 2010-04-08 0:33 ` Greg KH 2010-04-08 0:33 ` Greg KH 2010-04-09 0:01 ` Robert Hancock 2010-04-09 16:50 ` Sarah Sharp 2010-04-09 16:50 ` Sarah Sharp 2010-04-09 23:38 ` Robert Hancock 2010-04-09 23:38 ` Robert Hancock 2010-04-10 8:34 ` [alsa-devel] " Daniel Mack 2010-04-10 8:34 ` Daniel Mack 2010-04-10 17:02 ` [alsa-devel] " Robert Hancock 2010-04-12 18:56 ` Sarah Sharp 2010-04-12 20:39 ` Robert Hancock 2010-04-12 20:39 ` Robert Hancock 2010-04-12 20:58 ` Sarah Sharp 2010-04-12 11:17 ` [PATCH] USB: rename usb_buffer_alloc() and usb_buffer_free() Daniel Mack 2010-04-12 11:17 ` Daniel Mack 2010-04-13 18:16 ` Daniel Mack 2010-04-13 18:16 ` Daniel Mack 2010-04-13 19:27 ` Alan Stern 2010-04-13 19:27 ` Alan Stern 2010-04-13 20:26 ` Greg KH 2010-04-13 21:47 ` Daniel Mack 2010-04-13 21:47 ` Daniel Mack 2010-04-07 17:52 ` USB transfer_buffer allocations on 64bit systems Takashi Iwai 2010-04-07 17:52 ` Takashi Iwai 2010-04-07 15:46 ` Alan Stern 2010-04-07 15:46 ` Alan Stern 2010-04-08 6:12 ` Oliver Neukum 2010-04-08 16:59 ` Alan Stern 2010-04-08 16:59 ` Alan Stern 2010-04-08 21:24 ` Oliver Neukum 2010-04-08 22:20 ` Alan Stern 2010-04-08 22:20 ` Alan Stern 2010-04-09 6:04 ` Oliver Neukum 2010-04-09 14:41 ` Alan Stern [this message] 2010-04-09 14:41 ` Alan Stern 2010-04-09 14:50 ` Oliver Neukum 2010-04-09 14:50 ` Oliver Neukum 2010-04-09 15:15 ` Alan Stern 2010-04-09 15:15 ` Alan Stern 2010-04-09 20:51 ` Oliver Neukum 2010-04-09 20:51 ` Oliver Neukum 2010-04-09 21:21 ` Alan Stern 2010-04-09 21:21 ` Alan Stern 2010-04-07 16:54 ` Oliver Neukum 2010-04-07 16:54 ` Oliver Neukum 2010-04-07 17:00 ` Daniel Mack 2010-04-07 17:00 ` Daniel Mack 2010-04-07 23:55 ` Robert Hancock 2010-04-07 23:55 ` Robert Hancock 2010-04-08 2:10 ` Alan Stern 2010-04-08 2:10 ` Alan Stern 2010-04-08 7:30 ` Daniel Mack 2010-04-08 7:30 ` Daniel Mack 2010-04-08 16:57 ` Alan Stern 2010-04-08 16:57 ` Alan Stern 2010-04-08 17:17 ` Pedro Ribeiro 2010-04-08 18:17 ` Alan Stern 2010-04-08 23:13 ` Pedro Ribeiro 2010-04-08 23:13 ` Pedro Ribeiro 2010-04-09 16:01 ` Alan Stern 2010-04-09 16:01 ` Alan Stern 2010-04-09 18:09 ` Daniel Mack 2010-04-09 18:19 ` Pedro Ribeiro 2010-04-09 18:19 ` Pedro Ribeiro 2010-04-09 19:34 ` Alan Stern 2010-04-09 19:34 ` Alan Stern 2010-04-09 20:14 ` Daniel Mack 2010-04-09 20:14 ` Daniel Mack 2010-04-09 20:25 ` [LKML] " Konrad Rzeszutek Wilk 2010-04-09 20:25 ` Konrad Rzeszutek Wilk 2010-04-09 21:23 ` Alan Stern 2010-04-09 21:23 ` Alan Stern 2010-04-09 22:11 ` Robert Hancock 2010-04-12 10:48 ` Daniel Mack 2010-04-12 10:48 ` Daniel Mack 2010-04-12 12:06 ` Pedro Ribeiro 2010-04-10 12:49 ` Daniel Mack 2010-04-10 12:49 ` Daniel Mack 2010-04-10 13:21 ` Pedro Ribeiro 2010-04-10 13:21 ` Pedro Ribeiro 2010-04-12 8:59 ` Andi Kleen 2010-04-12 8:59 ` Andi Kleen 2010-04-12 11:14 ` Daniel Mack 2010-04-12 11:14 ` Daniel Mack 2010-04-12 11:53 ` Andi Kleen 2010-04-12 11:53 ` Andi Kleen 2010-04-12 12:11 ` Pedro Ribeiro 2010-04-12 12:12 ` Andi Kleen 2010-04-12 12:12 ` Andi Kleen 2010-04-12 12:32 ` Daniel Mack 2010-04-12 12:47 ` Andi Kleen 2010-04-12 12:54 ` Daniel Mack 2010-04-12 12:54 ` Daniel Mack 2010-04-12 15:43 ` Andi Kleen 2010-04-12 16:17 ` Alan Stern 2010-04-12 16:17 ` Alan Stern 2010-04-12 16:29 ` Andi Kleen 2010-04-12 16:57 ` Alan Stern 2010-04-12 16:57 ` Alan Stern 2010-04-12 17:15 ` Daniel Mack 2010-04-12 17:15 ` Daniel Mack 2010-04-12 17:22 ` Andi Kleen 2010-04-12 17:22 ` Andi Kleen 2010-04-12 17:56 ` Daniel Mack 2010-04-12 17:56 ` Daniel Mack 2010-04-12 17:52 ` [LKML] " Konrad Rzeszutek Wilk 2010-04-12 17:52 ` Konrad Rzeszutek Wilk 2010-04-13 18:22 ` Daniel Mack 2010-04-13 18:22 ` Daniel Mack 2010-04-13 23:46 ` Pedro Ribeiro 2010-04-13 23:46 ` Pedro Ribeiro 2010-04-14 10:09 ` Daniel Mack 2010-04-14 10:09 ` Daniel Mack 2010-04-14 10:47 ` Pedro Ribeiro 2010-04-14 11:02 ` Pedro Ribeiro 2010-04-14 13:18 ` [LKML] " Konrad Rzeszutek Wilk 2010-04-14 14:08 ` Alan Stern 2010-04-14 14:08 ` Alan Stern 2010-04-14 16:36 ` Daniel Mack 2010-04-14 16:36 ` Daniel Mack 2010-04-14 17:21 ` Pedro Ribeiro 2010-04-14 17:21 ` Pedro Ribeiro 2010-04-14 18:23 ` Alan Stern 2010-04-14 18:27 ` Pedro Ribeiro 2010-04-14 18:53 ` Alan Stern 2010-04-15 7:35 ` Daniel Mack 2010-04-15 7:35 ` Daniel Mack 2010-04-14 18:15 ` Alan Stern 2010-04-14 18:36 ` David Woodhouse 2010-04-14 21:12 ` Pedro Ribeiro 2010-04-14 22:25 ` Chris Wright 2010-04-14 22:56 ` Pedro Ribeiro 2010-04-14 23:37 ` Chris Wright 2010-04-15 1:20 ` Pedro Ribeiro 2010-04-15 15:20 ` Alan Stern 2010-04-20 0:16 ` Pedro Ribeiro 2010-05-07 7:48 ` Daniel Mack 2010-05-07 7:48 ` Daniel Mack 2010-05-07 9:47 ` [alsa-devel] " Clemens Ladisch 2010-05-07 9:47 ` Clemens Ladisch 2010-05-07 10:24 ` [alsa-devel] " Daniel Mack 2010-05-07 10:24 ` Daniel Mack 2010-05-07 14:51 ` [alsa-devel] " Alan Stern 2010-05-07 14:51 ` Alan Stern 2010-05-10 2:50 ` FUJITA Tomonori 2010-05-10 9:21 ` David Woodhouse 2010-05-10 9:21 ` David Woodhouse 2010-05-10 14:58 ` [alsa-devel] " Alan Stern 2010-05-10 14:58 ` Alan Stern 2010-05-11 1:06 ` FUJITA Tomonori 2010-05-11 1:06 ` FUJITA Tomonori 2010-05-11 14:00 ` Alan Stern 2010-05-11 14:00 ` Alan Stern 2010-05-11 14:22 ` FUJITA Tomonori 2010-05-11 14:24 ` Konrad Rzeszutek Wilk 2010-05-11 14:38 ` FUJITA Tomonori 2010-05-11 15:04 ` Alan Stern 2010-05-11 15:04 ` Alan Stern 2010-05-11 15:34 ` FUJITA Tomonori 2010-05-11 16:06 ` Alan Stern 2010-05-11 16:09 ` Daniel Mack 2010-05-11 16:48 ` Pedro Ribeiro 2010-05-11 17:10 ` Daniel Mack 2010-05-11 17:32 ` Pedro Ribeiro 2010-05-11 17:38 ` Daniel Mack 2010-05-12 23:50 ` Pedro Ribeiro 2010-05-13 9:36 ` Daniel Mack 2010-05-14 0:17 ` Pedro Ribeiro 2010-05-11 14:57 ` Alan Stern 2010-05-11 15:05 ` Daniel Mack 2010-05-10 14:31 ` Konrad Rzeszutek Wilk 2010-05-10 14:31 ` Konrad Rzeszutek Wilk 2010-05-07 11:42 ` Oliver Neukum 2010-05-07 11:42 ` Oliver Neukum 2010-05-07 11:47 ` Oliver Neukum 2010-05-07 11:47 ` Oliver Neukum 2010-05-07 11:58 ` Daniel Mack 2010-05-07 11:58 ` Daniel Mack 2010-05-07 14:45 ` [alsa-devel] " Alan Stern 2010-05-07 14:45 ` Alan Stern 2010-04-14 18:38 ` Chris Wright 2010-04-14 20:29 ` Alan Stern 2010-04-14 21:01 ` Konrad Rzeszutek Wilk 2010-04-14 21:12 ` Pedro Ribeiro 2010-04-15 1:50 ` Alan Stern
Reply instructions: You may reply publicly to this message via plain-text email using any one of the following methods: * Save the following mbox file, import it into your mail client, and reply-to-all from there: mbox Avoid top-posting and favor interleaved quoting: https://en.wikipedia.org/wiki/Posting_style#Interleaved_style * Reply using the --to, --cc, and --in-reply-to switches of git-send-email(1): git send-email \ --in-reply-to=Pine.LNX.4.44L0.1004091033150.1852-100000@iolanthe.rowland.org \ --to=stern@rowland.harvard.edu \ --cc=akpm@linux-foundation.org \ --cc=alsa-devel@alsa-project.org \ --cc=daniel@caiaq.de \ --cc=gregkh@suse.de \ --cc=linux-kernel@vger.kernel.org \ --cc=linux-usb@vger.kernel.org \ --cc=oliver@neukum.org \ --cc=pedrib@gmail.com \ /path/to/YOUR_REPLY https://kernel.org/pub/software/scm/git/docs/git-send-email.html * If your mail client supports setting the In-Reply-To header via mailto: links, try the mailto: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.