From: Matthew Dharm <mdharm-scsi@one-eyed-alien.net>
To: James Bottomley <James.Bottomley@steeleye.com>
Cc: Linus Torvalds <torvalds@transmeta.com>,
USB Developers <linux-usb-devel@lists.sourceforge.net>,
USB Storage List <usb-storage@one-eyed-alien.net>,
Linux SCSI list <linux-scsi@vger.kernel.org>
Subject: Re: PATCH: exclude certain commands from emulated SCSI hosts
Date: Sat, 5 Apr 2003 11:27:59 -0800 [thread overview]
Message-ID: <20030405112759.B4320@one-eyed-alien.net> (raw)
In-Reply-To: <1049556643.1762.16.camel@mulgrave>; from James.Bottomley@steeleye.com on Sat, Apr 05, 2003 at 09:30:41AM -0600
[-- Attachment #1: Type: text/plain, Size: 5834 bytes --]
I don't think this will work. I see two problems:
(1) You can't filter TEST_UNIT_READY (opcode 0). Not a big deal, but a
theoretical problem.
(2) We need to be able to filter at the originator. For example,
MODE_SENSE is perfectly fine to send to a CD-ROM, but not to a hard-disk.
We can't make that distinction with your code, unless we put the filtering
code not in queuecommand but in places like sd.c.... or we need to change
the filter to also take a device type.
Matt
On Sat, Apr 05, 2003 at 09:30:41AM -0600, James Bottomley wrote:
> On Mon, 2003-03-24 at 11:30, Matthew Dharm wrote:
> > On Mon, Mar 24, 2003 at 09:15:57AM -0600, James Bottomley wrote:
> > > OK, I can do this: A simple one with either a blacklist (reject these
> > > commands) or whitelist (only accept these commands) going by the first
> > > command byte OK?
> >
> > Well, you need to go by more than the first command byte -- ex. INQUIRY is
> > okay, unless length != 36 or EVPD.
> >
> > I think a blacklist is probably in order, but with a BIG COMMENT mentioning
> > that if someone adds new commands into the code paths they should at least
> > consider if they belong in the blacklist.
>
> OK, try the attached. There is no central blacklist, you must construct
> it on a per driver basis, so in the queuecommand of your driver you have
> a:
>
> static struct scsi_cmd_filter filter =
> { SCSI_FILTER_BLACKLIST ,
> { MODE_SENSE, SCSI_FILTER_INQUIRY_NOT36, 0 } };
>
> if(scsi_filter_cmd(SCp, &filter)) {
> SCp->scsi_done(SCp);
> return 0;
> }
>
> to use the filter (I think, I've compiled but not tested it).
>
> Let me know how it goes.
>
> James
>
> ===== drivers/scsi/scsi.h 1.68 vs edited =====
> --- 1.68/drivers/scsi/scsi.h Mon Mar 24 11:23:37 2003
> +++ edited/drivers/scsi/scsi.h Sat Apr 5 09:24:38 2003
> @@ -959,4 +959,39 @@
> extern int scsi_sysfs_register(void);
> extern void scsi_sysfs_unregister(void);
>
> +struct scsi_cmd_filter {
> + enum {
> + SCSI_FILTER_WHITELIST,
> + SCSI_FILTER_BLACKLIST
> + } type;
> + /* normal scsi commands are bytes, exceptions are words. The format
> + * of the exceptions is:
> + *
> + * Byte 15: invert the specific condition
> + * Byte 14-12: Reserved
> + * Byte 11-8: command opcode (must be > 0)
> + * Byte 7-0: Command specific */
> + unsigned short commands[];
> +};
> +
> +extern int scsi_filter_cmd(struct scsi_cmnd *, struct scsi_cmd_filter *);
> +
> +/* exception definitions for the filter */
> +#define SCSI_FILTER_INVERT 0x8000
> +
> +/* opcodes for the filter */
> +#define SCSI_FILTER_INQUIRY 0x0100
> +
> +/* useful filter commands */
> +#define SCSI_FILTER_INQUIRY_36 (SCSI_FILTER_INQUIRY | 36)
> +#define SCSI_FILTER_INQUIRY_NOT36 (SCSI_FILTER_INVERT | SCSI_FILTER_INQUIRY | 36)
> +
> +static inline unsigned short scsi_filter_opcode(unsigned short command) {
> + return command & 0x7f00;
> +}
> +
> +static inline unsigned char scsi_filter_data(unsigned short command) {
> + return (unsigned char)(command & 0x00ff);
> +}
> +
> #endif /* _SCSI_H */
> ===== drivers/scsi/scsi_lib.c 1.75 vs edited =====
> --- 1.75/drivers/scsi/scsi_lib.c Fri Mar 14 18:35:27 2003
> +++ edited/drivers/scsi/scsi_lib.c Sat Apr 5 09:27:35 2003
> @@ -1375,3 +1375,71 @@
> kmem_cache_destroy(sgp->slab);
> }
> }
> +
> +static inline int scsi_filter_exceptions(struct scsi_cmnd *cmd,
> + unsigned short command)
> +{
> + int found = 0;
> +
> + /* specials begin at 0x100 */
> + if(command < 0x100)
> + return 0;
> +
> + switch(scsi_filter_opcode(command)) {
> +
> + case SCSI_FILTER_INQUIRY:
> + if(cmd->cmnd[0] != INQUIRY)
> + return 0;
> + /* does the transfer length match the data */
> + found = (cmd->cmnd[4] == scsi_filter_data(command));
> + /* now check for inversion */
> + if(command & SCSI_FILTER_INVERT)
> + found = !found;
> + return found;
> + default:
> + /* unrecognized filter */
> + return 0;
> + }
> +}
> +
> +/**
> + * scsi_filter_cmd - Filter a given command against a list
> + * @cmd: command to be filtered.
> + * @filter: pointer to the filter containing the type (black/white list) and
> + * zero terminated list of commands to filter against (first byte only).
> + *
> + * Returns 0 if the filter passed successfully and the driver can continue
> + * processing the command or 1 if the filter failed and the command should
> + * be finished (via ->scsi_done). In the latter case, the command will have
> + * the sense fields filled in indicating the correct sense for an illegal
> + * request.
> + **/
> +int scsi_filter_cmd(struct scsi_cmnd *cmd, struct scsi_cmd_filter *filter)
> +{
> + int found = 0;
> + unsigned short *command;
> +
> + for(command = filter->commands; *command; command++) {
> + found = scsi_filter_exceptions(cmd, *command);
> + found += (cmd->cmnd[0] != *command);
> + if(found)
> + break;
> + }
> +
> + if((found && filter->type == SCSI_FILTER_WHITELIST) ||
> + (!found && filter->type == SCSI_FILTER_BLACKLIST))
> + return 0;
> +
> + /* fill in the Check Condition/Illegal request */
> + cmd->result = SAM_STAT_CHECK_CONDITION;
> + memset(cmd->sense_buffer, '\0', sizeof(cmd->sense_buffer));
> + cmd->sense_buffer[0] = 0xF0; /* valid, response code 0x70 */
> + cmd->sense_buffer[2] = ILLEGAL_REQUEST;
> +
> + /* ASC 0x20, ASCQ 0x00: Invalid command operation code */
> + cmd->sense_buffer[12] = 0x20;
> + cmd->sense_buffer[13] = 0x00;
> +
> + return 1;
> +}
> +
--
Matthew Dharm Home: mdharm-usb@one-eyed-alien.net
Maintainer, Linux USB Mass Storage Driver
IT KEEPS ASKING ME WHERE I WANT TO GO TODAY! I DONT WANT TO GO ANYWHERE!
-- Greg
User Friendly, 11/28/97
[-- Attachment #2: Type: application/pgp-signature, Size: 232 bytes --]
next prev parent reply other threads:[~2003-04-05 19:16 UTC|newest]
Thread overview: 45+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <20030322193046.A17056@one-eyed-alien.net>
[not found] ` <20030322193149.B17056@one-eyed-alien.net>
2003-03-23 3:37 ` PATCH: exclude certain commands from emulated SCSI hosts Matthew Dharm
2003-03-23 4:09 ` Linus Torvalds
2003-03-23 7:31 ` Matthew Dharm
2003-03-23 7:39 ` Linus Torvalds
2003-03-23 18:13 ` [usb-storage] " Matthew Dharm
2003-03-24 1:05 ` Douglas Gilbert
2003-03-24 1:26 ` James Bottomley
2003-03-24 1:37 ` Matthew Dharm
2003-03-24 1:39 ` James Bottomley
2003-03-24 7:04 ` Matthew Dharm
2003-03-24 15:15 ` James Bottomley
2003-03-24 16:29 ` Linus Torvalds
2003-03-24 16:43 ` James Bottomley
2003-03-24 16:52 ` Jens Axboe
2003-03-24 16:56 ` James Bottomley
2003-03-24 17:30 ` Matthew Dharm
2003-04-05 15:30 ` James Bottomley
2003-04-05 19:27 ` Matthew Dharm [this message]
2003-04-05 19:45 ` James Bottomley
2003-04-05 19:55 ` Matthew Dharm
2003-04-05 20:08 ` James Bottomley
2003-04-06 0:20 ` Matthew Dharm
2003-04-06 0:22 ` Matthew Dharm
2003-04-06 15:39 ` James Bottomley
2003-04-07 22:33 ` Patrick Mansfield
2003-04-07 23:14 ` James Bottomley
2003-04-08 0:51 ` Patrick Mansfield
2003-04-20 21:33 ` Matthew Dharm
2003-04-20 21:35 ` Matthew Dharm
2003-04-21 16:20 ` James Bottomley
2003-04-21 17:02 ` Matthew Dharm
2003-04-21 16:28 ` James Bottomley
2003-04-21 17:01 ` Matthew Dharm
2003-04-21 19:23 ` James Bottomley
2003-04-21 19:35 ` Matthew Dharm
2003-04-21 21:27 ` James Bottomley
2003-04-21 23:37 ` Matthew Dharm
2003-04-21 21:28 ` Patrick Mansfield
2003-04-21 23:45 ` Matthew Dharm
2003-03-24 1:58 ` Linus Torvalds
2003-03-24 6:58 ` Matthew Dharm
2003-04-22 17:37 [usb-storage] " James Bottomley
2003-04-22 18:13 ` Alan Stern
-- strict thread matches above, loose matches on Subject: below --
2003-04-22 19:30 Andries.Brouwer
2003-04-22 19:41 ` James Bottomley
2003-04-22 19:50 Andries.Brouwer
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=20030405112759.B4320@one-eyed-alien.net \
--to=mdharm-scsi@one-eyed-alien.net \
--cc=James.Bottomley@steeleye.com \
--cc=linux-scsi@vger.kernel.org \
--cc=linux-usb-devel@lists.sourceforge.net \
--cc=torvalds@transmeta.com \
--cc=usb-storage@one-eyed-alien.net \
/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