Thanks. This looks interesting. I like the way you provided a detailed
explanation of the motivation for the patch.
Do you agree that the offset should be a vlong so that it can support the same
large values that the size can support? An unsigned int sometimes isn't going
to be 64 bits.
On May 8, 2014, at 4:43 PM, Christoph Biedl
<sourceforge.b...@manchmal.in-ulm.de> wrote:
> Hello,
>
> the patch below implements a new option -o for vbladed, denoting the
> number of sectors that should be skipped at the beginning of the
> exported file. Default is 0 (no offset, export from beginning, current
> behaviour).
>
> Rationale:
>
> If the exported files are actually raw block devices on the target,
> something that might happen when using logical volumes for exports,
> several tools running on the target that assume block devices are
> for exclusive local access only might create confusion or even
> havoc. Most prominently:
>
> * os-prober of grub (might be disabled, though)
> Might mount any block device found that contain a file system.
> * (Linux) logical volume management
> Will report any physical volume if the export is used as such.
> * (Linux) blkid
> Reports any block device (minor impact, mostly information leakage)
>
> The overall problem is applications are not aware some block devices
> might be used from a remote client while there's no locking and that
> might be hard to implement
>
> Solutions tried but not considered helpful:
>
> * Filter out exported block devices
> Problems: The tools need the ability to filter; new tools might come
> into existance, requiring new action. Filter languages might be hard
> to use (Linux LVM is quite a nightmare here).
> * Optionally open block devices with O_EXCL
> Had no effect. Still exposes block devices where no vbladed is running
> yet.
>
> So the goal was to hide the actual content from these tools. One
> approach was to obscure the data using XOR, however it seemed easier
> to sacrifice one or more sectors (8 recommended for 4k sector hard
> drives) for a simple workaround.
>
> This is based on vblade-20, tests on Debian Linux included a simple
> write of each sector with a unique ID and later re-read in order to
> detect any re-ordering bugs, however everything worked as expected.
>
> Christoph
>
> --- a/aoe.c
> +++ b/aoe.c
> @@ -464,9 +464,10 @@ main(int argc, char **argv)
> int ch, omode = 0, readonly = 0;
>
> bufcnt = Bufcount;
> + offset = 0;
> setbuf(stdin, NULL);
> progname = *argv;
> - while ((ch = getopt(argc, argv, "b:dsrm:")) != -1) {
> + while ((ch = getopt(argc, argv, "b:dsrm:o:")) != -1) {
> switch (ch) {
> case 'b':
> bufcnt = atoi(optarg);
> @@ -485,6 +486,9 @@ main(int argc, char **argv)
> case 'm':
> setmask(optarg);
> break;
> + case 'o':
> + offset = atoi(optarg);
> + break;
> case '?':
> default:
> usage();
> @@ -505,6 +509,11 @@ main(int argc, char **argv)
> setserial(argv[3], shelf, slot);
> size = getsize(bfd);
> size /= 512;
> + if (size < offset) {
> + fprintf(stderr, "Offset %u too big - remaining size is
> negative!\n", offset);
> + exit(1);
> + }
> + size -= offset;
> ifname = argv[2];
> sfd = dial(ifname, bufcnt);
> getea(sfd, ifname, mac);
> diff --git a/ata.c b/ata.c
> index 4854f34..7c53b46 100644
> --- a/ata.c
> +++ b/ata.c
> @@ -155,12 +155,12 @@ atacmd(Ataregs *p, uchar *dp, int ndp, int payload) //
> do the ata cmd
> return 0;
> }
> if (p->cmd == 0x20 || p->cmd == 0x24)
> - n = getsec(bfd, dp, lba, p->sectors);
> + n = getsec(bfd, dp, lba+offset, p->sectors);
> else {
> // packet should be big enough to contain the data
> if (payload < 512 * p->sectors)
> return -1;
> - n = putsec(bfd, dp, lba, p->sectors);
> + n = putsec(bfd, dp, lba+offset, p->sectors);
> }
> n /= 512;
> if (n != p->sectors) {
> diff --git a/dat.h b/dat.h
> index 064ab1e..f3a93d8 100644
> --- a/dat.h
> +++ b/dat.h
> @@ -169,6 +169,7 @@ uchar mac[6];
> int bfd; // block file descriptor
> int sfd; // socket file descriptor
> vlong size; // size of vblade
> +unsigned offset;
> char *progname;
> char serial[Nserial+1];
>
> diff --git a/vblade.8 b/vblade.8
> index 2cd2b2c..3d8ff13 100644
> --- a/vblade.8
> +++ b/vblade.8
> @@ -55,6 +55,10 @@ The -r flag restricts the export of the device to be
> read-only.
> The -m flag takes an argument, a comma separated list of MAC addresses
> permitted access to the vblade. A MAC address can be specified in upper
> or lower case, with or without colons.
> +.TP
> +\fB-o\fP
> +The -o flag takes an argument, the number of sectors that should be
> +skipped at the beginning of filename.
> .SH EXAMPLE
> In this example, the root user on a host named
> .I nai
> ------------------------------------------------------------------------------
> Is your legacy SCM system holding you back? Join Perforce May 7 to find out:
> • 3 signs your SCM is hindering your productivity
> • Requirements for releasing software faster
> • Expert tips and advice for migrating your SCM now
> http://p.sf.net/sfu/perforce_______________________________________________
> Aoetools-discuss mailing list
> Aoetools-discuss@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/aoetools-discuss
------------------------------------------------------------------------------
Is your legacy SCM system holding you back? Join Perforce May 7 to find out:
• 3 signs your SCM is hindering your productivity
• Requirements for releasing software faster
• Expert tips and advice for migrating your SCM now
http://p.sf.net/sfu/perforce
_______________________________________________
Aoetools-discuss mailing list
Aoetools-discuss@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/aoetools-discuss
