lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Date:	Sat, 09 Jun 2007 14:13:20 +0400
From:	Sergei Shtylyov <sshtylyov@...mvista.com>
To:	Bartlomiej Zolnierkiewicz <bzolnier@...il.com>
Cc:	Geller Sandor <wildy@...ra.hos.u-szeged.hu>,
	Andrew Morton <akpm@...ux-foundation.org>,
	linux-kernel@...r.kernel.org, linux-ide@...r.kernel.org
Subject: Re: HPT374 IDE problem with 2.6.21.* kernels

Bartlomiej Zolnierkiewicz wrote:

>>>>>>Sergei, do we need to disallow UDMA6 completely on HPT734 or
>>>>>>is it only an issue with some problematic devices (=> blacklist)?

>>>>   Note that I didn't change what the old code was doing in this regard -- 
>>>>although the HPT374 spec does *not* say that UDMA6 is supported, it had been 
>>>>enabled. What have *really* changed for HPT374 was:

>>>>- in 2.6.20-rc1, the driver switched to using the actual 33 MHz timing table
>>>>  instead of the old one, matching 50 MHz (and so, severely underclocked);

>>>>- in 2.6.2-rc1, the driver switched from 33 MHz PCI to 66 MHz DPLL clock.

>>>>   Disallowing UDMA6 would clock the chip with 50 MHz DPLL, howewer, the 

>>>I felt inspired by this explanation (thanks!) and took a look at
>>>hpt374-opensource-v2.10 vendor driver.  Here is something interesting:
>>
>>>glbdata.c:
>>
>>>...
>>>#ifdef CLOCK_66MHZ
>>>ULONG setting370_66[] = {
>>>        0xd029d5e,  0xd029d26,  0xc829ca6,  0xc829c84,  0xc829c62,
>>>        0x2c829d2c, 0x2c829c66, 0x2c829c62,
>>>        0x1c829c62, 0x1c9a9c62, 0x1c929c62, 0x1c8e9c62, 0x1c8a9c62,
>>>        0x1c8a9c62/*0x1cae9c62*/, 0x1c869c62, 0x1c869c62,
>>>};
>>>...

>>>hpt366.c:

>>>...
>>>static u32 sixty_six_base_hpt37x[] = {
>>>        /* XFER_UDMA_6 */       0x1c869c62,
>>>        /* XFER_UDMA_5 */       0x1cae9c62,     /* 0x1c8a9c62 */
>>>...

>>>So we are using Dual ATA Clock for UDMA5 whereas vendor driver doesn't

>>    This is so in all other HPT drivers (and HPT371N datasheet has the same 
>>figures -- this chip is the only one supporting UDMA6 and having the default 
>>DPLL clock > 50 MHz).  Note that it means that there's no actual UDMA5 since 
>>the timing exactly matches that one used for UDMA4.

>>>(the only other mode which uses Dual ATA Clock, in both drivers, is rarely
>>>used UDMA3).

>>    And UDMA4 with 50 MHz clock.

>>>Thanks to this UDMA cycle time should be equal 22.5ns instead of 30ns
>>>(spec defines it at 16.8ns, ide_timings[] uses 20ns) when using 66 MHz DPLL
>>>clock.  In theory everything should play nice but the data manual for HPT374

>>    And it does -- on other chips.

> My beautiful theory failed... Oh, well... ;)

    Sigh, if we only knew why HPT decided that UDMA5 timings should be the 
same as UDMA4 -- probably they had some reason...

>>>contains weird note that Dual ATA Clock is meant to implement ATA100 read
>>>and write at different clocks (there is no more explanation to this).

>>    That's the thing that keeps me confused in the other datasheets too -- 
>>from my interpretation of their timing figures it seemed to control 2x ATA 
>>clock multipler. HPT370 datasheet just gives different timings and SCR2 values 
>>for reads/writes in UDMA5 (I've disabled this mode on HPT370 from which the 
>>read performance only gained -- not sure if it makes sense to restore the old 
>>clock turnaround hack).

    It used to clock the writes from DPLL in UDMA5, and clock UDMA5 reads and 
all other modes from PCI... And the result was dog slow reads in UDMA5 which 
UDMA4 was beating by about 8 MB/s... Maybe that's why UDMA4 timings were used 
for UDMA5 in later chips by HPT -- but the real UDMA5 yielded faster transfer 
speeds than UDMA4 for those chips...

>>>Geller reported that the problems started after migrating from 2.6.20.7 to
>>>2.6.21.1 (the affected disks are using UDMA5) and at the same time the driver
>>>switched from 33 MHz PCI to 66 MHz DPLL clock.  Also the issue is completely
>>>fixed by using 50 MHz DPLL clock (UDMA5 timing for 50 MHz DPLL clock is
>>>0x12848242 so UDMA cycle time equals 20ns and is smaller than the one
>>>obtained using 66 MHz DPLL clock).

>>>It all makes me wonder whether it is really safe to use Dual ATA Clock for
>>>UDMA5 and whether we should just be using "the offical" timing instead...

>>    Not sure. I had no problems with this on the HPT371N/302 and 371N was 
>>clocked by 66 MHz DPLL from the start (its default clock is 75 MHz however).

    I meant to say 77... :-)

>>    I'm still holding to my hypothesis that HPT374 simply can't tolerate 66 
>>MHz DPLL clock, and the UDMA5 timing figures that you've cited seem to prove that.
>>    I'm going to post a patch today -- how about completely prohibiting UDMA6 
>>on HPT374?

> Sounds fine, in case somebody misses it we can introduce something like
> hpt374_allow_66mhz_dpll module parameter...

    Don't think anybody will miss it. Anyway, chip spec doesn't say that it's 
supported.

> Thanks,
> Bart

MBR, Sergei
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Powered by blists - more mailing lists