linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Boris Brezillon <boris.brezillon@free-electrons.com>
To: Masahiro Yamada <yamada.masahiro@socionext.com>
Cc: linux-mtd@lists.infradead.org, devicetree@vger.kernel.org,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
	Marek Vasut <marek.vasut@gmail.com>,
	Brian Norris <computersforpeace@gmail.com>,
	Richard Weinberger <richard@nod.at>,
	David Woodhouse <dwmw2@infradead.org>,
	Cyrille Pitchen <cyrille.pitchen@atmel.com>,
	Rob Herring <robh+dt@kernel.org>,
	Mark Rutland <mark.rutland@arm.com>,
	Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Subject: Re: [PATCH 00/39] mtd: nand: denali: 2nd round of Denali NAND IP patch bomb
Date: Mon, 13 Mar 2017 12:33:22 +0100	[thread overview]
Message-ID: <20170313123322.06475e5c@bbrezillon> (raw)
In-Reply-To: <CAK7LNAQxTSJT3szu46pQfBXGmenAexMWo8GyjmHsnXVmht4mOA@mail.gmail.com>

Hi Masahiro,

On Fri, 10 Mar 2017 20:00:03 +0900
Masahiro Yamada <yamada.masahiro@socionext.com> wrote:

> Hi Boris,
> 
> I am almost getting v2 done,
> and now I am testing it.
> 
> I am having one problem.  Please teach me.
> 
> 
> 2016-11-30 17:17 GMT+09:00 Boris Brezillon <boris.brezillon@free-electrons.com>:
> >> [2]
> >> Remove driver-internal bounce buffer.
> >> The current Denali driver allocate DMA_BIDIRECTIONAL buffer
> >> to use it as a driver-internal bounce buffer.
> >>
> >> The hardware transfer page data into the bounce buffer,
> >> then CPU copies from the bounce buffer to a given buf (and oob_poi).
> >> This is not efficient.
> >>
> >> So, I want to set NAND_USE_BOUNCE_BUFFER flag
> >> and do dma_map_single directly for a given buffer.  
> >
> > Sounds good. Be careful though, when you use the generic bounce buffer
> > interface you might have to clear the page cache info (->pagebuf = -1).  
> 
> 
> Instead of memcpy() of the whole page,
> I am trying to use dma_map_single()  in ecc->read_page() / ecc->write_page().
> This will allow direct transfer between the buffer and the device by DMA.
> 
> But, this does not work for Denali if use_bufpoi is set in nand_do_read_ops().
> 
> 
> In the following code in nand_scan_tail(),
> 
>         if (!(chip->options & NAND_OWN_BUFFERS)) {
>                 nbuf = kzalloc(sizeof(*nbuf) + mtd->writesize
>                                 + mtd->oobsize * 3, GFP_KERNEL);
>                 if (!nbuf)
>                         return -ENOMEM;
>                 nbuf->ecccalc = (uint8_t *)(nbuf + 1);
>                 nbuf->ecccode = nbuf->ecccalc + mtd->oobsize;
>                 nbuf->databuf = nbuf->ecccode + mtd->oobsize;
> 
>                 chip->buffers = nbuf;
> 
> 
> chip->buffers->databuf has no guarantee for DMA'able alignment.
> (actually it has unwanted offset 0xc because sizeof(*nbuf) == 0xc on
> 32bit systems)

Well, I think the DMA alignment requirement is a platform/controller
specific (some controllers are fine with this 32bits alignment), but I
get your point.

> 
> If we could change the code as follows,
> 
>                 nbuf->ecccalc = kmalloc(mtd->oobsize, GFP_KERNEL);
>                 nbuf->ecccode = kmalloc(mtd->oobsize, GFP_KERNEL);
>                 nbuf->databuf = kmalloc(mtd->writesize + mtd->oobsize,
> GFP_KERNEL);
> 
> chip->buffers->databuf would have DMA'able alignment in most cases
> without NAND_OWN_BUFFERS.  (but, I am not sure if this is a good idea)

I'm fine with this change. I don't know what are the guarantees in term
of alignment when you use kmalloc, but I guess the size you're
allocating (writesize + oobsize) kind of guarantees that the alignment
is rather big (because the SLAB caches are organized by power-of-2
chunk sizes, and for allocations >PAGE_SIZE the page allocator will be
used).

> 
> 
> So, the idea of NAND_OWN_BUFFERS is that
> drivers should allocate own buffers if they need to perform DMA-mapping
> in read_page(), write_page(), right?

Right.

> 
> 
> However, "git grep NAND_OWN_BUFFERS" shows
> cafe_nand.c is the only driver that does so.
> 
> On the other hand, "git grep dma_map_single" has more hits,
> i.e. some drivers perform dma_map_single() for read/write without
> NAND_OWN_BUFFERS.
> 
> I have no idea how they are working.

Probably because the controllers and/or DMA engines have no alignment
constraints.

Anyway, the change you're proposing is rather simple, so go ahead.

Regards,

Boris

  reply	other threads:[~2017-03-13 11:33 UTC|newest]

Thread overview: 94+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-11-26 18:05 [PATCH 00/39] mtd: nand: denali: 2nd round of Denali NAND IP patch bomb Masahiro Yamada
2016-11-26 18:05 ` [PATCH 01/39] mtd: nand: allow to set only one of ECC size and ECC strength from DT Masahiro Yamada
2016-11-27 13:58   ` Boris Brezillon
2016-11-26 18:05 ` [PATCH 02/39] mtd: nand: denali: remove unused CONFIG option and macros Masahiro Yamada
2016-11-26 18:05 ` [PATCH 03/39] mtd: nand: denali: remove redundant define of BANK(x) Masahiro Yamada
2016-11-26 18:05 ` [PATCH 04/39] mtd: nand: denali: remove more unused struct members Masahiro Yamada
2016-11-27 15:12   ` Boris Brezillon
2016-11-30  7:16     ` Masahiro Yamada
2016-11-30  7:47       ` Boris Brezillon
2016-11-26 18:05 ` [PATCH 05/39] mtd: nand: denali: fix comment of denali_nand_info::flash_mem Masahiro Yamada
2016-11-26 18:05 ` [PATCH 06/39] mtd: nand: denali: fix write_oob_data() function Masahiro Yamada
2016-11-26 18:05 ` [PATCH 07/39] mtd: nand: denali: transfer OOB only when oob_required is set Masahiro Yamada
2016-11-26 18:05 ` [PATCH 08/39] mtd: nand: denali: introduce capability flag Masahiro Yamada
2016-11-26 18:05 ` [PATCH 09/39] mtd: nand: denali: fix erased page check code Masahiro Yamada
2016-11-27 15:21   ` Boris Brezillon
2016-12-02  4:33     ` Masahiro Yamada
2016-12-02  7:57       ` Boris Brezillon
2016-11-26 18:05 ` [PATCH 10/39] mtd: nand: denali: remove redundant if conditional of erased_check Masahiro Yamada
2016-11-26 18:05 ` [PATCH 11/39] mtd: nand: denali: increment ecc_stats.failed by one per error Masahiro Yamada
2016-11-26 18:05 ` [PATCH 12/39] mtd: nand: denali: return 0 for uncorrectable ECC error Masahiro Yamada
2016-11-26 18:05 ` [PATCH 13/39] mtd: nand: denali: increment ecc_stats->corrected Masahiro Yamada
2016-11-27 15:31   ` Boris Brezillon
2016-12-02  4:28     ` Masahiro Yamada
2016-11-26 18:06 ` [PATCH 14/39] mtd: nand: denali: replace uint{8/16/32}_t with u{8/16/32} Masahiro Yamada
2016-11-26 18:06 ` [PATCH 15/39] mtd: nand: denali: improve readability of handle_ecc() Masahiro Yamada
2016-11-27 15:34   ` Boris Brezillon
2016-11-27 15:42   ` Boris Brezillon
2016-12-02  4:26     ` Masahiro Yamada
2016-12-02  7:55       ` Boris Brezillon
2016-11-26 18:06 ` [PATCH 16/39] mtd: nand: denali: rename handle_ecc() to denali_sw_ecc_fixup() Masahiro Yamada
2016-11-26 18:06 ` [PATCH 17/39] mtd: nand: denali: support HW_ECC_FIXUP capability Masahiro Yamada
2016-11-27 16:09   ` Boris Brezillon
2016-11-30  6:20     ` Masahiro Yamada
2016-11-30  7:51       ` Boris Brezillon
2016-11-26 18:06 ` [PATCH 18/39] mtd: nand: denali: move denali_read_page_raw() above denali_read_page() Masahiro Yamada
2016-11-27 16:10   ` Boris Brezillon
2016-11-30  6:13     ` Masahiro Yamada
2016-11-26 18:06 ` [PATCH 19/39] mtd: nand: denali: perform erased check against raw transferred page Masahiro Yamada
2016-11-27 16:12   ` Boris Brezillon
2016-11-30  5:36     ` Masahiro Yamada
2016-11-30  8:00       ` Boris Brezillon
2016-11-26 18:06 ` [PATCH 20/39] mtd: nand: denali_dt: enable HW_ECC_FIXUP capability for DT platform Masahiro Yamada
2016-11-27 16:14   ` Boris Brezillon
2016-11-30  6:14     ` Masahiro Yamada
2016-11-26 18:06 ` [PATCH 21/39] mtd: nand: denali: support 64bit capable DMA engine Masahiro Yamada
2016-11-27 16:16   ` Boris Brezillon
2016-11-30  7:11     ` Masahiro Yamada
2016-11-30  7:17       ` Boris Brezillon
2016-11-26 18:06 ` [PATCH 22/39] mtd: nand: denali_dt: remove dma-mask DT property Masahiro Yamada
2016-12-01 15:56   ` Rob Herring
2016-11-26 18:06 ` [PATCH 23/39] mtd: nand: denali_dt: use pdev instead of ofdev for platform_device Masahiro Yamada
2016-11-26 18:06 ` [PATCH 24/39] mtd: nand: denali: add NEW_N_BANKS_FORMAT capability Masahiro Yamada
2016-11-26 18:06 ` [PATCH 25/39] mtd: nand: denali: use nand_chip to hold frequently accessed data Masahiro Yamada
2016-11-26 18:06 ` [PATCH 26/39] mtd: nand: denali: call nand_set_flash_node() to set DT node Masahiro Yamada
2016-11-26 18:06 ` [PATCH 27/39] mtd: nand: denali: do not set mtd->name Masahiro Yamada
2016-11-26 18:06 ` [PATCH 28/39] mtd: nand: denali: move multi NAND fixup code to a helper function Masahiro Yamada
2016-11-27 16:24   ` Boris Brezillon
2016-11-30  6:09     ` Masahiro Yamada
2016-11-30  8:07       ` Boris Brezillon
2016-11-26 18:06 ` [PATCH 29/39] mtd: nand: denali: refactor multi NAND fixup code in more generic way Masahiro Yamada
2016-11-26 18:06 ` [PATCH 30/39] mtd: nand: denali: set DEVICES_CONNECTED 1 if not set Masahiro Yamada
2016-11-26 18:06 ` [PATCH 31/39] mtd: nand: denali: remove meaningless writes to read-only registers Masahiro Yamada
2016-11-26 18:06 ` [PATCH 32/39] mtd: nand: denali: remove unnecessary writes to ECC_CORRECTION Masahiro Yamada
2016-11-26 18:06 ` [PATCH 33/39] mtd: nand: denali: support 1024 byte ECC step size Masahiro Yamada
2016-12-01 15:58   ` Rob Herring
2016-11-26 18:06 ` [PATCH 34/39] mtd: nand: denali: fix the condition for 15 bit ECC strength Masahiro Yamada
2016-11-26 18:06 ` [PATCH 35/39] mtd: nand: denali: calculate ecc.strength and ecc.bytes generically Masahiro Yamada
2016-11-26 18:06 ` [PATCH 36/39] mtd: nand: denali: allow to use SoC-specific ECC strength Masahiro Yamada
2016-11-26 18:06 ` [PATCH 37/39] mtd: nand: denali: support "nand-ecc-strength" DT property Masahiro Yamada
2016-12-01 15:59   ` Rob Herring
2016-11-26 18:06 ` [PATCH 38/39] mtd: nand: denali: remove Toshiba, Hynix specific fixup code Masahiro Yamada
2016-11-27 16:28   ` Boris Brezillon
2016-11-26 18:06 ` [PATCH 39/39] mtd: nand: denali_dt: add compatible strings for UniPhier SoC variants Masahiro Yamada
2016-12-01 16:05   ` Rob Herring
2016-12-02  2:54     ` Masahiro Yamada
2016-12-02 16:26       ` Rob Herring
2016-12-03  2:41         ` Masahiro Yamada
2016-12-03  2:49           ` Marek Vasut
2016-12-03 22:08             ` Dinh Nguyen
2016-12-05  3:30               ` Masahiro Yamada
2016-12-05  3:44                 ` Marek Vasut
2016-12-05  4:10                   ` Masahiro Yamada
2016-12-05  4:22                     ` Marek Vasut
2016-12-05 20:51                       ` Dinh Nguyen
2016-12-05 21:29                         ` Marek Vasut
2016-12-05 22:31                           ` Dinh Nguyen
2016-11-27 15:04 ` [PATCH 00/39] mtd: nand: denali: 2nd round of Denali NAND IP patch bomb Boris Brezillon
2016-11-30  8:02   ` Masahiro Yamada
2016-11-30  8:17     ` Boris Brezillon
2016-12-01  9:15       ` Masahiro Yamada
2017-03-10 11:00       ` Masahiro Yamada
2017-03-13 11:33         ` Boris Brezillon [this message]
2016-11-30  8:13   ` Masahiro Yamada
2016-11-27 16:31 ` Boris Brezillon

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=20170313123322.06475e5c@bbrezillon \
    --to=boris.brezillon@free-electrons.com \
    --cc=andriy.shevchenko@linux.intel.com \
    --cc=computersforpeace@gmail.com \
    --cc=cyrille.pitchen@atmel.com \
    --cc=devicetree@vger.kernel.org \
    --cc=dwmw2@infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mtd@lists.infradead.org \
    --cc=marek.vasut@gmail.com \
    --cc=mark.rutland@arm.com \
    --cc=richard@nod.at \
    --cc=robh+dt@kernel.org \
    --cc=yamada.masahiro@socionext.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: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).