linux-kernel - Re: [PATCH v1] mtd: rawnand: Add Macronix NAND read retry support

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] [thread-next>] [day] [month] [year] [list]

Date:   Fri, 10 May 2019 15:53:40 +0200
From:   Miquel Raynal <miquel.raynal@...tlin.com>
To:     Thomas Petazzoni <thomas.petazzoni@...tlin.com>
Cc:     Mason Yang <masonccyang@...c.com.tw>, bbrezillon@...nel.org,
        marek.vasut@...il.com, linux-kernel@...r.kernel.org,
        richard@....at, dwmw2@...radead.org, computersforpeace@...il.com,
        linux-mtd@...ts.infradead.org, juliensu@...c.com.tw
Subject: Re: [PATCH v1] mtd: rawnand: Add Macronix NAND read retry support

Hi Mason,

Thomas Petazzoni <thomas.petazzoni@...tlin.com> wrote on Fri, 10 May
2019 15:37:04 +0200:

> Hello,
> 
> Some purely cosmetic suggestions below.
> 
> On Fri, 10 May 2019 15:41:02 +0800
> Mason Yang <masonccyang@...c.com.tw> wrote:
> 
> > +	if (ret)
> > +		pr_err("set feature failed to read retry moded:%d\n", mode);    
> 
> I don't know what is the policy in the MTD/NAND subsystem, but
> shouldn't you be using dev_err() instead of pr_err() here to have a
> nice prefix for the message ?
> 
> 		dev_err(&nand_to_mtd(chip)->dev, "set feature ..", mode);

Indeed. You can even dereference an mtd_info object first, then use
mtd->dev.

> 
> > +static void macronix_nand_onfi_init(struct nand_chip *chip)
> > +{
> > +	struct nand_parameters *p = &chip->parameters;
> > +
> > +	if (p->onfi) {  
> 
> Change to:
> 
> 	if (!p->onfi)
> 		return;
> 
> This way the rest of the function can save one level of indentation.
> 
> > +		struct nand_onfi_vendor_macronix *mxic =
> > +				(void *)p->onfi->vendor;
> > +
> > +		if (mxic->reliability_func & MACRONIX_READ_RETRY_BIT) {  
> 
> Change to:
> 
> 	if (mxic->reliability_func & MACRONIX_READ_RETRY_BIT == 0)
> 		return;
> 
> And the rest of the function can save one level of indentation.
> 
> > +			chip->read_retries = MACRONIX_READ_RETRY_MODE + 1;
> > +			chip->setup_read_retry =
> > +				 macronix_nand_setup_read_retry;
> > +			if (p->supports_set_get_features) {
> > +				set_bit(ONFI_FEATURE_ADDR_READ_RETRY,
> > +					p->set_feature_list);
> > +				set_bit(ONFI_FEATURE_ADDR_READ_RETRY,
> > +					p->get_feature_list);
> > +			}  
> 
> Which will require less wrapping in those lines that are already at the
> third indentation level.
> 
> To me, it is also more logical: we exclude the cases we are not
> interested in and return early, and then if we are still in the case we
> are interested, we handle it.

I definitely agree with these cosmetic changes.

Thanks,
Miquèl