From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mail.karo-electronics.de ([81.173.242.67]) by merlin.infradead.org with esmtps (Exim 4.80.1 #2 (Red Hat Linux)) id 1W5zw0-0002IO-Om for linux-mtd@lists.infradead.org; Wed, 22 Jan 2014 15:41:45 +0000 Date: Wed, 22 Jan 2014 16:41:15 +0100 From: Lothar =?UTF-8?B?V2HDn21hbm4=?= To: Akinobu Mita Subject: Re: [PATCH] mtd: mtd_oobtest: fix verify errors due to incorrect use of prandom_bytes_state() Message-ID: <20140122164115.211a819d@ipc1.ka-ro> In-Reply-To: References: <1389608739-10945-1-git-send-email-LW@KARO-electronics.de> <20140122150959.78cf2827@ipc1.ka-ro> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Cc: Artem Bityutskiy , LKML , Huang Shijie , "linux-mtd@lists.infradead.org" , Andrew Morton , Brian Norris , David Woodhouse List-Id: Linux MTD discussion mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Hi, Akinobu Mita wrote: > 2014/1/22 Lothar Wa=C3=9Fmann : > > Hi, > > > > Is anyone taking care of this? > > > > Lothar Wa=C3=9Fmann wrote: > >> When using prandom_bytes_state() it is critical to use the same block > >> size in all invocations that are to produce the same random sequence. > >> Otherwise the state of the PRNG will be out of sync if the blocksize > >> is not divisible by 4. > >> This leads to bogus verification errors in several tests which use > >> different block sizes to initialize the buffer for writing and > >> comparison. > >> > >> Signed-off-by: Lothar Wa=C3=9Fmann > >> --- > >> drivers/mtd/tests/oobtest.c | 14 ++++++++++++-- > >> 1 files changed, 12 insertions(+), 2 deletions(-) > >> > >> diff --git a/drivers/mtd/tests/oobtest.c b/drivers/mtd/tests/oobtest.c > >> index 2e9e2d1..72c7359 100644 > >> --- a/drivers/mtd/tests/oobtest.c > >> +++ b/drivers/mtd/tests/oobtest.c > >> @@ -213,8 +213,15 @@ static int verify_eraseblock_in_one_go(int ebnum) > >> int err =3D 0; > >> loff_t addr =3D ebnum * mtd->erasesize; > >> size_t len =3D mtd->ecclayout->oobavail * pgcnt; > >> + int i; > >> + > >> + for (i =3D 0; i < pgcnt; i++) > >> + prandom_bytes_state(&rnd_state, &writebuf[i * use_len], > >> + use_len); > >> + if (len % use_len) > >> + prandom_bytes_state(&rnd_state, &writebuf[i * use_len], > >> + len % use_len); > >> > >> - prandom_bytes_state(&rnd_state, writebuf, len); > >> ops.mode =3D MTD_OPS_AUTO_OOB; > >> ops.len =3D 0; > >> ops.retlen =3D 0; >=20 > I would rather fix the use of prandom_bytes_state() in write_eraseblock() > than fix in verify_eraseblock_in_one_go(). >=20 Why and how? write_whole_device() (which calls write_eraseblock()) is used multiple times with different verification methods (all blocks in one go or each block individually). If prandom_state_bytes() in write_eraseblock() would be changed, that function would have to know, how the block are going to be checked lateron to know how to set up the writebuffer. Dto. for mtd_write_oob() Lothar Wa=C3=9Fmann --=20 ___________________________________________________________ Ka-Ro electronics GmbH | Pascalstra=C3=9Fe 22 | D - 52076 Aachen Phone: +49 2408 1402-0 | Fax: +49 2408 1402-10 Gesch=C3=A4ftsf=C3=BChrer: Matthias Kaussen Handelsregistereintrag: Amtsgericht Aachen, HRB 4996 www.karo-electronics.de | info@karo-electronics.de ___________________________________________________________