linux-kernel - [PATCH] block/rnbd: Fix the maximum clt_device_id value in init

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

Date:   Sun, 27 Mar 2022 13:06:30 +0200
From:   Christophe JAILLET <christophe.jaillet@...adoo.fr>
To:     "Md. Haris Iqbal" <haris.iqbal@...os.com>,
        Jack Wang <jinpu.wang@...os.com>, Jens Axboe <axboe@...nel.dk>,
        Bart Van Assche <bvanassche@....org>,
        Jason Gunthorpe <jgg@...pe.ca>,
        Danil Kipnis <danil.kipnis@...ud.ionos.com>
Cc:     linux-kernel@...r.kernel.org, kernel-janitors@...r.kernel.org,
        Christophe JAILLET <christophe.jaillet@...adoo.fr>,
        Jack Wang <jinpu.wang@...ud.ionos.com>,
        linux-block@...r.kernel.org
Subject: [PATCH] block/rnbd: Fix the maximum clt_device_id value in init_dev()

ida_alloc_range(..., min, max, ...) returns values from min to max,
inclusive.

So, '1 << (MINORBITS - RNBD_PART_BITS)' is a valid value for ret, which is
then saved in 'dev->clt_device_id'.

This value is used in rnbd_client_setup_device() and passed to
rnbd_clt_setup_gen_disk().
There we have:
    dev->gd->first_minor	= idx << RNBD_PART_BITS

So a possible value for 'gd->first_minor' is '1 << MINORBITS'

This is an issue because:
    rnbd_clt_setup_gen_disk()
    --> add_disk(dev->gd)
      --> device_add_disk(NULL, disk, NULL)

And there we have:
   ddev->devt = MKDEV(disk->major, disk->first_minor);

So, should 'gd->first_minor' be '1 << MINORBITS', MKDEV() would overflow.

Fixes: f7a7a5c228d4 ("block/rnbd: client: main functionality")
Signed-off-by: Christophe JAILLET <christophe.jaillet@...adoo.fr>
---
#define MKDEV(ma,mi)	(((ma) << MINORBITS) | (mi))

This patch is completely speculative.

I think that:
	if (disk->first_minor + disk->minors > MINORMASK + 1)
		return -EINVAL;
in device_add_disk() handles this corner case.
Anyway, if I'm correct, handling the error earlier can't hurt (at least I
guess so :)).

Signed-off-by: Christophe JAILLET <christophe.jaillet@...adoo.fr>
---
 drivers/block/rnbd/rnbd-clt.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/block/rnbd/rnbd-clt.c b/drivers/block/rnbd/rnbd-clt.c
index b66e8840b94b..db900c3786a3 100644
--- a/drivers/block/rnbd/rnbd-clt.c
+++ b/drivers/block/rnbd/rnbd-clt.c
@@ -1454,7 +1454,7 @@ static struct rnbd_clt_dev *init_dev(struct rnbd_clt_session *sess,
 		goto out_alloc;
 	}
 
-	ret = ida_alloc_max(&index_ida, 1 << (MINORBITS - RNBD_PART_BITS),
+	ret = ida_alloc_max(&index_ida, (1 << (MINORBITS - RNBD_PART_BITS)) - 1,
 			    GFP_KERNEL);
 	if (ret < 0) {
 		pr_err("Failed to initialize device '%s' from session %s, allocating idr failed, err: %d\n",
-- 
2.32.0