linux-kernel - Re: [RFC PATCH] mtd: lpddr: Fix bad logic in print_drs

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 linux-hardening linux-cve-announce 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:   Mon, 27 Apr 2020 13:09:43 -0500
From:   "Gustavo A. R. Silva" <gustavo@...eddedor.com>
To:     Joe Perches <joe@...ches.com>,
        Miquel Raynal <miquel.raynal@...tlin.com>
Cc:     Richard Weinberger <richard@....at>,
        Vignesh Raghavendra <vigneshr@...com>,
        linux-mtd@...ts.infradead.org,
        linux-kernel <linux-kernel@...r.kernel.org>
Subject: Re: [RFC PATCH] mtd: lpddr: Fix bad logic in print_drs_error

Hi,

It seems that this fix should be tagged for -stable.

I can create a three-patch series for this (as Miquel suggested), starting
with the patch that fix the wrong condition, so it can be ported to
-stable, separately.

I'll include you Reported-by, Joe.

Thanks
--
Gustavo

On 4/27/20 12:33, Joe Perches wrote:
> On Mon, 2020-04-27 at 15:29 +0200, Miquel Raynal wrote:
>> Hi Joe,
> 
> Hello Miquel.
> 
>> Joe Perches <joe@...ches.com> wrote on Mon, 30 Mar 2020 12:56:59 -0700:
>>
>>> Update logic for broken test.
>>> Use a more common logging style.
>>>
>>> Miscellanea:
>>>
>>> o Coalesce formats
>>>
>>> Signed-off-by: Joe Perches <joe@...ches.com>
>>> ---
>>>
>>> Found by inspection of include files using printk.
>>>
>>> It appears the logic in this function is broken for the
>>> consecutive tests of
>>>
>>> 	if (prog_status & 0x3)
>>> 		...
>>> 	else if (prog_status & 0x2)
>>> 		...
>>> 	else (prog_status & 0x1)
>>> 		...
>>>
>>> Likely the first test should be
>>>
>>> 	if ((prog_status & 0x3) == 0x3)
>>
>> I had a hard time understanding the patch just with the commit log, I
>> think the above text is as important.
>>
>> In fact, would you mind doing the printk->pr_notice in a first patch,
>> and fix the wrong condition in a separate change?
>>
>>> And this function is only used in drivers/mtd/lpddr/lpddr_cmds.c
>>> perhaps it should be moved there.
>>
>> Agreed, do not hesitate to move the function as suggested in a third
>> patch.
> 
> You are the maintainer here no?
> 
> I think you (or perhaps the author Alexey Korolev but he hasn't
> been active in a decade) should be doing all this.
> 
> I just identified the logic defect.
> 
>>>  include/linux/mtd/pfow.h | 31 ++++++++++++++-----------------
>>>  1 file changed, 14 insertions(+), 17 deletions(-)
>>>
>>> diff --git a/include/linux/mtd/pfow.h b/include/linux/mtd/pfow.h
>>> index 122f343..1c08e75 100644
>>> --- a/include/linux/mtd/pfow.h
>>> +++ b/include/linux/mtd/pfow.h
>>> @@ -127,31 +127,28 @@ static inline void print_drs_error(unsigned dsr)
>>>  	int prog_status = (dsr & DSR_RPS) >> 8;
>>>  
>>>  	if (!(dsr & DSR_AVAILABLE))
>>> -		printk(KERN_NOTICE"DSR.15: (0) Device not Available\n");
>>> -	if (prog_status & 0x03)
>>> -		printk(KERN_NOTICE"DSR.9,8: (11) Attempt to program invalid "
>>> -						"half with 41h command\n");
>>> +		pr_notice("DSR.15: (0) Device not Available\n");
>>> +
>>> +	if ((prog_status & 0x03) == 0x03)
>>> +		pr_notice("DSR.9,8: (11) Attempt to program invalid half with 41h command\n");
>>>  	else if (prog_status & 0x02)
>>> -		printk(KERN_NOTICE"DSR.9,8: (10) Object Mode Program attempt "
>>> -					"in region with Control Mode data\n");
>>> +		pr_notice("DSR.9,8: (10) Object Mode Program attempt in region with Control Mode data\n");
>>>  	else if (prog_status &  0x01)
>>> -		printk(KERN_NOTICE"DSR.9,8: (01) Program attempt in region "
>>> -						"with Object Mode data\n");
>>> +		pr_notice("DSR.9,8: (01) Program attempt in region with Object Mode data\n");
>>> +
>>>  	if (!(dsr & DSR_READY_STATUS))
>>> -		printk(KERN_NOTICE"DSR.7: (0) Device is Busy\n");
>>> +		pr_notice("DSR.7: (0) Device is Busy\n");
>>>  	if (dsr & DSR_ESS)
>>> -		printk(KERN_NOTICE"DSR.6: (1) Erase Suspended\n");
>>> +		pr_notice("DSR.6: (1) Erase Suspended\n");
>>>  	if (dsr & DSR_ERASE_STATUS)
>>> -		printk(KERN_NOTICE"DSR.5: (1) Erase/Blank check error\n");
>>> +		pr_notice("DSR.5: (1) Erase/Blank check error\n");
>>>  	if (dsr & DSR_PROGRAM_STATUS)
>>> -		printk(KERN_NOTICE"DSR.4: (1) Program Error\n");
>>> +		pr_notice("DSR.4: (1) Program Error\n");
>>>  	if (dsr & DSR_VPPS)
>>> -		printk(KERN_NOTICE"DSR.3: (1) Vpp low detect, operation "
>>> -					"aborted\n");
>>> +		pr_notice("DSR.3: (1) Vpp low detect, operation aborted\n");
>>>  	if (dsr & DSR_PSS)
>>> -		printk(KERN_NOTICE"DSR.2: (1) Program suspended\n");
>>> +		pr_notice("DSR.2: (1) Program suspended\n");
>>>  	if (dsr & DSR_DPS)
>>> -		printk(KERN_NOTICE"DSR.1: (1) Aborted Erase/Program attempt "
>>> -					"on locked block\n");
>>> +		pr_notice("DSR.1: (1) Aborted Erase/Program attempt on locked block\n");
>>>  }
>>>  #endif /* __LINUX_MTD_PFOW_H */
>>>
>>>
>>
>> Thanks,
>> Miquèl
>