From: Alan Stern <stern@rowland.harvard.edu>
To: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
Cc: Felipe Balbi <balbi@kernel.org>,
Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
linux-usb@vger.kernel.org, linux-kernel@vger.kernel.org,
Petr Nechaev <petr.nechaev@cogentembedded.com>
Subject: Re: [PATCH] usb: gadget: storage: add support for media larger than 2T
Date: Tue, 14 Sep 2021 11:13:29 -0400 [thread overview]
Message-ID: <20210914151329.GD155245@rowland.harvard.edu> (raw)
In-Reply-To: <20210914052728.23369-1-nikita.yoush@cogentembedded.com>
On Tue, Sep 14, 2021 at 08:27:28AM +0300, Nikita Yushchenko wrote:
> This adds support for READ_CAPACITY(16), READ(16) and WRITE(16)
> commands, and fixes READ_CAPACITY command to return 0xffffffff if
> media size does not fit in 32 bits.
>
> This makes f_mass_storage to export a 16T disk array correctly.
>
> Signed-off-by: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
> ---
> drivers/usb/gadget/function/f_mass_storage.c | 70 ++++++++++++++++++--
> 1 file changed, 63 insertions(+), 7 deletions(-)
>
> diff --git a/drivers/usb/gadget/function/f_mass_storage.c b/drivers/usb/gadget/function/f_mass_storage.c
> index 7c96c4665178..b79737c19750 100644
> --- a/drivers/usb/gadget/function/f_mass_storage.c
> +++ b/drivers/usb/gadget/function/f_mass_storage.c
> @@ -619,7 +619,7 @@ static int sleep_thread(struct fsg_common *common, bool can_freeze,
> static int do_read(struct fsg_common *common)
> {
> struct fsg_lun *curlun = common->curlun;
> - u32 lba;
> + u64 lba;
> struct fsg_buffhd *bh;
> int rc;
> u32 amount_left;
> @@ -634,7 +634,10 @@ static int do_read(struct fsg_common *common)
> if (common->cmnd[0] == READ_6)
> lba = get_unaligned_be24(&common->cmnd[1]);
> else {
> - lba = get_unaligned_be32(&common->cmnd[2]);
> + if (common->cmnd[0] == READ_16)
> + lba = get_unaligned_be64(&common->cmnd[2]);
> + else
It would be good to put a comment here:
else /* READ_10 or READ_12 */
Same for the WRITE_10/WRITE_12 case below.
> + lba = get_unaligned_be32(&common->cmnd[2]);
>
> /*
> * We allow DPO (Disable Page Out = don't save data in the
> @@ -747,7 +750,7 @@ static int do_read(struct fsg_common *common)
> static int do_write(struct fsg_common *common)
> {
> struct fsg_lun *curlun = common->curlun;
> - u32 lba;
> + u64 lba;
> struct fsg_buffhd *bh;
> int get_some_more;
> u32 amount_left_to_req, amount_left_to_write;
> @@ -771,7 +774,10 @@ static int do_write(struct fsg_common *common)
> if (common->cmnd[0] == WRITE_6)
> lba = get_unaligned_be24(&common->cmnd[1]);
> else {
> - lba = get_unaligned_be32(&common->cmnd[2]);
> + if (common->cmnd[0] == WRITE_16)
> + lba = get_unaligned_be64(&common->cmnd[2]);
> + else
> + lba = get_unaligned_be32(&common->cmnd[2]);
>
> /*
> * We allow DPO (Disable Page Out = don't save data in the
> @@ -1146,6 +1152,7 @@ static int do_read_capacity(struct fsg_common *common, struct fsg_buffhd *bh)
> u32 lba = get_unaligned_be32(&common->cmnd[2]);
> int pmi = common->cmnd[8];
> u8 *buf = (u8 *)bh->buf;
> + u32 mlb;
The Linux kernel does not suffer from a shortage of letters; you don't
have to struggle to conserve them. This variable should be named
max_lba or something similar. (Note: To readers in the US, "mlb" stands
for "Major League Baseball"!)
>
> /* Check the PMI and LBA fields */
> if (pmi > 1 || (pmi == 0 && lba != 0)) {
> @@ -1153,12 +1160,28 @@ static int do_read_capacity(struct fsg_common *common, struct fsg_buffhd *bh)
> return -EINVAL;
> }
>
> - put_unaligned_be32(curlun->num_sectors - 1, &buf[0]);
> - /* Max logical block */
> - put_unaligned_be32(curlun->blksize, &buf[4]);/* Block length */
> + if (curlun->num_sectors < 0x100000000ULL)
> + mlb = curlun->num_sectors - 1;
> + else
> + mlb = 0xffffffff;
> + put_unaligned_be32(mlb, &buf[0]); /* Max logical block */
> + put_unaligned_be32(curlun->blksize, &buf[4]); /* Block length */
> return 8;
> }
>
> +static int do_read_capacity_16(struct fsg_common *common, struct fsg_buffhd *bh)
> +{
> + struct fsg_lun *curlun = common->curlun;
> + u8 *buf = (u8 *)bh->buf;
> +
> + put_unaligned_be64(curlun->num_sectors - 1, &buf[0]);
> + /* Max logical block */
> + put_unaligned_be32(curlun->blksize, &buf[8]); /* Block length */
> + /* It is safe to keep other fields zeroed */
Blank line preceding the comment, please.
> + memset(&buf[12], 0, 32 - 12);
> + return 32;
> +}
> +
> static int do_read_header(struct fsg_common *common, struct fsg_buffhd *bh)
> {
> struct fsg_lun *curlun = common->curlun;
> @@ -1905,6 +1928,17 @@ static int do_scsi_command(struct fsg_common *common)
> reply = do_read(common);
> break;
>
> + case READ_16:
> + common->data_size_from_cmnd =
> + get_unaligned_be32(&common->cmnd[10]);
> + reply = check_command_size_in_blocks(common, 16,
> + DATA_DIR_TO_HOST,
> + (1<<1) | (0xff<<2) | (0xf<<10), 1,
> + "READ(16)");
> + if (reply == 0)
> + reply = do_read(common);
> + break;
> +
> case READ_CAPACITY:
> common->data_size_from_cmnd = 8;
> reply = check_command(common, 10, DATA_DIR_TO_HOST,
> @@ -1914,6 +1948,17 @@ static int do_scsi_command(struct fsg_common *common)
> reply = do_read_capacity(common, bh);
> break;
>
> + case SERVICE_ACTION_IN_16:
I realize that this is the code for READ_CAPACITY_16, but the commands
should be kept in alphabetical order by symbol name. So this belongs
lower down, after REQUEST_SENSE.
> + if ((common->cmnd[1] & 0x1f) != SAI_READ_CAPACITY_16)
> + goto unknown_cmnd;
> + common->data_size_from_cmnd = 32;
This should be get_unaligned_be32(&common->cmnd[10]).
> + reply = check_command(common, 14, DATA_DIR_TO_HOST,
16, not 14.
> + (1<<1) | (1<<13), 1,
Do you want to disallow nonzero values for the LBA field? Then
this should be (1<<1) | (0xf<<10) | (1<<14). If you also want to
disallow nonzero values for the PMI bit, omit the (1<<14) part.
> + "READ CAPACITY(16)");
> + if (reply == 0)
> + reply = do_read_capacity_16(common, bh);
> + break;
> +
> case READ_HEADER:
> if (!common->curlun || !common->curlun->cdrom)
> goto unknown_cmnd;
> @@ -2028,6 +2073,17 @@ static int do_scsi_command(struct fsg_common *common)
> reply = do_write(common);
> break;
>
> + case WRITE_16:
> + common->data_size_from_cmnd =
> + get_unaligned_be32(&common->cmnd[10]);
> + reply = check_command_size_in_blocks(common, 16,
> + DATA_DIR_FROM_HOST,
> + (1<<1) | (0xff<<2) | (0xf<<10), 1,
> + "WRITE(16)");
> + if (reply == 0)
> + reply = do_write(common);
> + break;
> +
> /*
> * Some mandatory commands that we recognize but don't implement.
> * They don't mean much in this setting. It's left as an exercise
Please resubmit the patch with these changes.
Alan Stern
next prev parent reply other threads:[~2021-09-14 15:13 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-09-14 5:27 Nikita Yushchenko
2021-09-14 15:13 ` Alan Stern [this message]
2021-09-14 20:09 ` [PATCH v2] " Nikita Yushchenko
2021-09-14 20:32 ` Alan Stern
2021-09-21 14:29 ` Greg Kroah-Hartman
2021-09-21 14:59 ` [PATCH v3] " Nikita Yushchenko
2021-10-05 11:07 ` Greg Kroah-Hartman
2021-10-05 12:07 ` Felipe Balbi
2021-10-05 14:19 ` Alan Stern
2021-10-10 13:08 ` Greg Kroah-Hartman
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20210914151329.GD155245@rowland.harvard.edu \
--to=stern@rowland.harvard.edu \
--cc=balbi@kernel.org \
--cc=gregkh@linuxfoundation.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-usb@vger.kernel.org \
--cc=nikita.yoush@cogentembedded.com \
--cc=petr.nechaev@cogentembedded.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®