public inbox for u-boot@lists.denx.de
 help / color / mirror / Atom feed
From: Lukasz Majewski <l.majewski@samsung.com>
To: u-boot@lists.denx.de
Subject: [U-Boot] [PATCH 3/3] dfu: Introduction of the "dfu_checksum_method" env variable for checksum method setting
Date: Mon, 31 Mar 2014 14:04:38 +0200	[thread overview]
Message-ID: <20140331140438.65f4e5d9@amdc2363> (raw)
In-Reply-To: <545BA3D1-2228-46F0-B224-134B9C57CDF6@antoniou-consulting.com>

Hi Pantelis,

> Hi Lukasz,
> 
> On Mar 31, 2014, at 11:48 AM, Lukasz Majewski wrote:
> 
> > Up till now the CRC32 of received data was calculated
> > unconditionally. The standard crc32 implementation causes long
> > delays when large images were uploaded.
> > 
> > The "dfu_checksum_method" environment variable gives the
> > opportunity to enable on demand (when e.g. debugging) the crc32
> > calculation. It can be done without need to recompile the u-boot
> > binary.
> > 
> > By default the crc32 is not calculated.
> > 
> > Tests results:
> > 400 MiB ums.img file
> > With 		crc32 calculation: 65 sec [avg 6.29 MB/s]
> > Without 		crc32 calculation: 25 sec [avg 16.17 MB/s]
> > 
> 
> That's interesting; I'm surprised that there's so much difference.
> Can we get some info about the environment? I.e. board, whether cache
> is enabled etc.


Board Exynos4412 - Trats2. Cache L1 enabled, cache L2 disabled.

Crc32 is calculated for 32 MiB chunks of data in the received buffer.
I'm using the standard software crc32 u-boot's implementation. It is
the same as the one for perl-crc32 debian package.


> 
> The crc table is per byte and I guess lookups maybe expensive.

It seems so. Moreover the Exynos4412 has HW crypto engine which
supports SHA1, MD5 and other algorithms. Unfortunately it doesn't
provide speedup for crc32.


> 
> Regards
> 
> -- Pantelis
> 
> 
> > Signed-off-by: Lukasz Majewski <l.majewski@samsung.com>
> > ---
> > drivers/dfu/dfu.c |   34 ++++++++++++++++++++++++++++++----
> > include/dfu.h     |    5 +++++
> > 2 files changed, 35 insertions(+), 4 deletions(-)
> > 
> > diff --git a/drivers/dfu/dfu.c b/drivers/dfu/dfu.c
> > index e15f959..5d50b47 100644
> > --- a/drivers/dfu/dfu.c
> > +++ b/drivers/dfu/dfu.c
> > @@ -20,6 +20,7 @@ static bool dfu_reset_request;
> > static LIST_HEAD(dfu_list);
> > static int dfu_alt_num;
> > static int alt_num_count;
> > +static int dfu_checksum_method;
> > 
> > bool dfu_reset(void)
> > {
> > @@ -99,6 +100,23 @@ unsigned char *dfu_get_buf(void)
> > 	return dfu_buf;
> > }
> > 
> > +static int dfu_get_checksum_method(void)
> > +{
> > +	char *s;
> > +
> > +	s = getenv("dfu_checksum_method");
> > +	if (!s)
> > +		return DFU_NO_CHECKSUM;
> > +
> > +	if (!strcmp(s, "crc32")) {
> > +		debug("%s: DFU checksum method: %s\n", __func__,
> > s);
> > +		return DFU_CRC32;
> > +	} else {
> > +		error("DFU checksum method: %s not supported!\n",
> > s);
> > +		return -EINVAL;
> > +	}
> > +}
> > +
> > static int dfu_write_buffer_drain(struct dfu_entity *dfu)
> > {
> > 	long w_size;
> > @@ -109,8 +127,8 @@ static int dfu_write_buffer_drain(struct
> > dfu_entity *dfu) if (w_size == 0)
> > 		return 0;
> > 
> > -	/* update CRC32 */
> > -	dfu->crc = crc32(dfu->crc, dfu->i_buf_start, w_size);
> > +	if (dfu_checksum_method == DFU_CRC32)
> > +		dfu->crc = crc32(dfu->crc, dfu->i_buf_start,
> > w_size);
> > 
> > 	ret = dfu->write_medium(dfu, dfu->offset, dfu->i_buf_start,
> > &w_size); if (ret)
> > @@ -234,7 +252,8 @@ static int dfu_read_buffer_fill(struct
> > dfu_entity *dfu, void *buf, int size) /* consume */
> > 		if (chunk > 0) {
> > 			memcpy(buf, dfu->i_buf, chunk);
> > -			dfu->crc = crc32(dfu->crc, buf, chunk);
> > +			if (dfu_checksum_method == DFU_CRC32)
> > +				dfu->crc = crc32(dfu->crc, buf,
> > chunk); dfu->i_buf += chunk;
> > 			dfu->b_left -= chunk;
> > 			dfu->r_left -= chunk;
> > @@ -318,7 +337,9 @@ int dfu_read(struct dfu_entity *dfu, void *buf,
> > int size, int blk_seq_num) }
> > 
> > 	if (ret < size) {
> > -		debug("%s: %s CRC32: 0x%x\n", __func__, dfu->name,
> > dfu->crc);
> > +		if (dfu_checksum_method == DFU_CRC32)
> > +			debug("%s: %s CRC32: 0x%x\n", __func__,
> > dfu->name,
> > +			      dfu->crc);
> > 		puts("\nUPLOAD ... done\nCtrl+C to exit ...\n");
> > 
> > 		dfu_free_buf();
> > @@ -393,6 +414,11 @@ int dfu_config_entities(char *env, char
> > *interface, int num) dfu_alt_num = dfu_find_alt_num(env);
> > 	debug("%s: dfu_alt_num=%d\n", __func__, dfu_alt_num);
> > 
> > +	ret = dfu_get_checksum_method();
> > +	if (ret < 0)
> > +		return ret;
> > +	dfu_checksum_method = ret;
> > +
> > 	dfu = calloc(sizeof(*dfu), dfu_alt_num);
> > 	if (!dfu)
> > 		return -1;
> > diff --git a/include/dfu.h b/include/dfu.h
> > index 751f0fd..855d6dc 100644
> > --- a/include/dfu.h
> > +++ b/include/dfu.h
> > @@ -37,6 +37,11 @@ enum dfu_op {
> > 	DFU_OP_WRITE,
> > };
> > 
> > +enum dfu_checksum {
> > +	DFU_NO_CHECKSUM = 0,
> > +	DFU_CRC32,
> > +};
> > +
> > #define DFU_NOT_SUPPORTED -1
> > 
> > struct mmc_internal_data {
> > -- 
> > 1.7.10.4
> > 



-- 
Best regards,

Lukasz Majewski

Samsung R&D Institute Poland (SRPOL) | Linux Platform Group

  reply	other threads:[~2014-03-31 12:04 UTC|newest]

Thread overview: 51+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-03-31  8:48 [U-Boot] [PATCH 0/3] dfu: Several enhancements for dfu subsystem Lukasz Majewski
2014-03-31  8:48 ` [U-Boot] [PATCH 1/3] dfu: mmc: Provide support for eMMC boot partition access Lukasz Majewski
2014-03-31  8:59   ` Marek Vasut
2014-03-31  9:14     ` Lukasz Majewski
2014-05-09 14:58   ` [U-Boot] [PATCH v2] " Lukasz Majewski
2014-05-14 22:24     ` Marek Vasut
2014-03-31  8:48 ` [U-Boot] [PATCH 2/3] dfu: add static alt num count in dfu_config_entities() Lukasz Majewski
2014-03-31  9:01   ` Marek Vasut
2014-03-31  9:15     ` Lukasz Majewski
2014-04-01  6:47       ` Przemyslaw Marczak
2014-04-01  6:49         ` Marek Vasut
2014-04-01  7:45           ` Lukasz Majewski
2014-03-31  8:48 ` [U-Boot] [PATCH 3/3] dfu: Introduction of the "dfu_checksum_method" env variable for checksum method setting Lukasz Majewski
2014-03-31  9:04   ` Marek Vasut
2014-03-31  9:24     ` Lukasz Majewski
2014-03-31  9:29       ` Marek Vasut
2014-03-31  9:49         ` Lukasz Majewski
2014-03-31 11:19   ` Pantelis Antoniou
2014-03-31 12:04     ` Lukasz Majewski [this message]
2014-03-31 12:10       ` Pantelis Antoniou
2014-03-31 12:16       ` Pantelis Antoniou
2014-03-31 18:05   ` Tom Rini
2014-03-31 18:15     ` Marek Vasut
2014-03-31 18:26       ` Tom Rini
2014-03-31 20:44     ` Lukasz Majewski
2014-03-31 21:04       ` Tom Rini
2014-04-01  9:05         ` Lukasz Majewski
2014-03-31 21:44       ` Tormod Volden
2014-04-01  9:00         ` Lukasz Majewski
2014-04-01  9:15           ` Stefan Schmidt
2014-04-01 11:31             ` Lukasz Majewski
2014-05-05 13:16   ` [U-Boot] [PATCH v2] " Lukasz Majewski
2014-05-05 17:47     ` Marek Vasut
2014-05-08 12:27     ` [U-Boot] [PATCH v3] dfu: Introduction of the "dfu_hash_algo" " Lukasz Majewski
2014-05-08 13:07       ` Marek Vasut
2014-05-09  4:27       ` Wolfgang Denk
2014-05-09  6:52         ` Lukasz Majewski
2014-05-09  8:31           ` Wolfgang Denk
2014-05-09  9:54             ` Lukasz Majewski
2014-05-12 14:45             ` Tom Rini
2014-05-15  7:09               ` Lukasz Majewski
2014-05-15  9:27                 ` Heiko Schocher
2014-05-15 11:19                 ` Wolfgang Denk
2014-05-15 13:43                   ` Lukasz Majewski
2014-05-15 14:07                     ` Wolfgang Denk
2014-05-16  6:08                       ` Lukasz Majewski
2014-05-16  8:58                     ` Lukasz Majewski
2014-05-19 14:02                       ` Heiko Schocher
2014-05-20 17:22                         ` Lukasz Majewski
2014-05-22  9:46                         ` Lukasz Majewski
2014-05-12  8:43     ` [U-Boot] [PATCH v4] " Lukasz Majewski

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=20140331140438.65f4e5d9@amdc2363 \
    --to=l.majewski@samsung.com \
    --cc=u-boot@lists.denx.de \
    /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