RE: [PATCH v2] usb: gadget: f_mass_storage: Make CD-ROM emulation works with Windows OS

From: Neal Liu <neal_liu@aspeedtech.com>
To: Alan Stern <stern@rowland.harvard.edu>
Cc: Felipe Balbi <balbi@kernel.org>,
	Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	Roger Quadros <roger.quadros@nokia.com>,
	Miaohe Lin <linmiaohe@huawei.com>,
	Wesley Cheng <quic_wcheng@quicinc.com>,
	"Eric W . Biederman" <ebiederm@xmission.com>,
	Matthew Wilcox <willy@infradead.org>,
	Nikita Yushchenko <nikita.yoush@cogentembedded.com>,
	Cai Huoqing <cai.huoqing@linux.dev>,
	"linux-usb@vger.kernel.org" <linux-usb@vger.kernel.org>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
	BMC-SW <BMC-SW@aspeedtech.com>
Subject: RE: [PATCH v2] usb: gadget: f_mass_storage: Make CD-ROM emulation works with Windows OS
Date: Tue, 28 Jun 2022 01:57:49 +0000	[thread overview]
Message-ID: <HK0PR06MB3202C5E9147EA7E7B91C6B0380B89@HK0PR06MB3202.apcprd06.prod.outlook.com> (raw)
In-Reply-To: <YrnEli0fdOERpIXe@rowland.harvard.edu>

> On Mon, Jun 27, 2022 at 02:56:40AM +0000, Neal Liu wrote:
> > > On Fri, Jun 24, 2022 at 10:19:16AM +0800, Neal Liu wrote:
> > > > Add read TOC with format 1 to support CD-ROM emulation with
> > > > Windows OS.
> > > > This patch is tested on Windows OS Server 2019.
> > > >
> > > > Signed-off-by: Neal Liu <neal_liu@aspeedtech.com>
> > > > ---
> > > >  drivers/usb/gadget/function/f_mass_storage.c | 5 +++--
> > > >  1 file changed, 3 insertions(+), 2 deletions(-)
> > > >
> > > > diff --git a/drivers/usb/gadget/function/f_mass_storage.c
> > > > b/drivers/usb/gadget/function/f_mass_storage.c
> > > > index 3a77bca0ebe1..9edf76c22605 100644
> > > > --- a/drivers/usb/gadget/function/f_mass_storage.c
> > > > +++ b/drivers/usb/gadget/function/f_mass_storage.c
> > > > @@ -1209,7 +1209,8 @@ static int do_read_toc(struct fsg_common
> > > > *common, struct fsg_buffhd *bh)
> > > >
> > > >  	switch (format) {
> > > >  	case 0:
> > > > -		/* Formatted TOC */
> > > > +	case 1:
> > > > +		/* Formatted TOC, Session info */
> > >
> > > This comment is a little misleading.  0 is formatted TOC, and 1 is
> > > multi-session info.  The way you wrote it, it looks like the comment
> > > applies to both formats.  Do it like this:
> > >
> > > 	case 0:		/* Formatted TOC */
> > > 	case 1:		/* Multi-session info */
> > >
> >
> > Okay, looks more clear. I'll fix it in next patch.
> >
> > > >  		len = 4 + 2*8;		/* 4 byte header + 2 descriptors */
> > > >  		memset(buf, 0, len);
> > > >  		buf[1] = len - 2;	/* TOC Length excludes length field */
> > >
> > > It looks like you ignored one of the problems I pointed out in my earlier
> email:
> > >
> > > > When format is 1, the driver is supposed to ignore the start_track
> > > > value.  Your patch does not do this.
> > >
> > > Please fix this.
> > >
> >
> > Do you mean ignore the "start_track > 1" above?
> 
> Yes, that's what I mean.
> 
> > Do this check only for format 0? Or format 0 & 2?
> 
> According to the MMC spec, the Track/Session Number is ignored for format 1.
> For format 0 it is valid as a Track Number, and for format 2 it is valid as a
> Session Number.  Therefore it should be checked for formats 0 and 2 but not
> for format 1.
> 
> Alan Stern

Got it. I'll fix this in next patch, thanks for pointing out.