* [PATCH] scsi: qla2xxx: Fix flex array member not at end
@ 2026-07-14 15:22 Jesse Taube
2026-07-15 16:16 ` Kees Cook
0 siblings, 1 reply; 3+ messages in thread
From: Jesse Taube @ 2026-07-14 15:22 UTC (permalink / raw)
To: linux-kernel
Cc: linux-scsi, Nilesh Javali, GR-QLogic-Storage-Upstream,
James E.J. Bottomley, Martin K. Petersen, Gustavo A. R. Silva,
Kees Cook, Christoph Hellwig, John Meneghini, Chris Leech,
Jesse Taube, Jesse Taube
In qla_edif_bsg.h: `struct fc_bsg_reply` and `struct fc_bsg_request`
have flexible array members, thus they must be the last member of
the parent structure. Contininging in the effort to add
`-Wflex-array-member-not-at-end`, move the structs to the end of
the parent structures, `struct qla_bsg_auth_els_reply` and
`struct qla_bsg_auth_els_request `.
Suggested-by: John Meneghini <jmeneghi@redhat.com>
Signed-off-by: Jesse Taube <jtaubepe@redhat.com>
---
drivers/scsi/qla2xxx/qla_edif_bsg.h | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/scsi/qla2xxx/qla_edif_bsg.h b/drivers/scsi/qla2xxx/qla_edif_bsg.h
index 514c265ba86e..bcc35d149c1f 100644
--- a/drivers/scsi/qla2xxx/qla_edif_bsg.h
+++ b/drivers/scsi/qla2xxx/qla_edif_bsg.h
@@ -36,16 +36,16 @@ struct extra_auth_els {
} __packed;
struct qla_bsg_auth_els_request {
- struct fc_bsg_request r;
struct extra_auth_els e;
+ struct fc_bsg_request r;
};
struct qla_bsg_auth_els_reply {
- struct fc_bsg_reply r;
uint32_t rx_xchg_address;
uint8_t version;
uint8_t pad[VND_CMD_PAD_SIZE];
uint8_t reserved[VND_CMD_APP_RESERVED_SIZE];
+ struct fc_bsg_reply r;
};
struct app_id {
--
2.54.0
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH] scsi: qla2xxx: Fix flex array member not at end
2026-07-14 15:22 [PATCH] scsi: qla2xxx: Fix flex array member not at end Jesse Taube
@ 2026-07-15 16:16 ` Kees Cook
2026-07-17 15:45 ` Jesse Taube
0 siblings, 1 reply; 3+ messages in thread
From: Kees Cook @ 2026-07-15 16:16 UTC (permalink / raw)
To: Jesse Taube
Cc: linux-kernel, linux-scsi, Nilesh Javali,
GR-QLogic-Storage-Upstream, James E.J. Bottomley,
Martin K. Petersen, Gustavo A. R. Silva, Christoph Hellwig,
John Meneghini, Chris Leech, Jesse Taube
On Tue, Jul 14, 2026 at 11:22:38AM -0400, Jesse Taube wrote:
> In qla_edif_bsg.h: `struct fc_bsg_reply` and `struct fc_bsg_request`
> have flexible array members, thus they must be the last member of
> the parent structure. Contininging in the effort to add
> `-Wflex-array-member-not-at-end`, move the structs to the end of
> the parent structures, `struct qla_bsg_auth_els_reply` and
> `struct qla_bsg_auth_els_request `.
>
> Suggested-by: John Meneghini <jmeneghi@redhat.com>
> Signed-off-by: Jesse Taube <jtaubepe@redhat.com>
It feels like this structure is part of the hardware spec? (and therefore
cannot be re-ordered.) I'm not sure at all, though, but it looks like
it's part of request serialization:
$ git grep 'qla_bsg_auth_els_request' | grep '>request'
drivers/scsi/qla2xxx/qla_edif.c: (struct qla_bsg_auth_els_request *)bsg_job->request;
drivers/scsi/qla2xxx/qla_edif.c: (struct qla_bsg_auth_els_request *)bsg_job->request;
drivers/scsi/qla2xxx/qla_edif.c: (struct qla_bsg_auth_els_request *)bsg_job->request;
drivers/scsi/qla2xxx/qla_isr.c: (struct qla_bsg_auth_els_request *)bsg_job->request;
-Kees
> ---
> drivers/scsi/qla2xxx/qla_edif_bsg.h | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/scsi/qla2xxx/qla_edif_bsg.h b/drivers/scsi/qla2xxx/qla_edif_bsg.h
> index 514c265ba86e..bcc35d149c1f 100644
> --- a/drivers/scsi/qla2xxx/qla_edif_bsg.h
> +++ b/drivers/scsi/qla2xxx/qla_edif_bsg.h
> @@ -36,16 +36,16 @@ struct extra_auth_els {
> } __packed;
>
> struct qla_bsg_auth_els_request {
> - struct fc_bsg_request r;
> struct extra_auth_els e;
> + struct fc_bsg_request r;
> };
>
> struct qla_bsg_auth_els_reply {
> - struct fc_bsg_reply r;
> uint32_t rx_xchg_address;
> uint8_t version;
> uint8_t pad[VND_CMD_PAD_SIZE];
> uint8_t reserved[VND_CMD_APP_RESERVED_SIZE];
> + struct fc_bsg_reply r;
> };
>
> struct app_id {
> --
> 2.54.0
>
--
Kees Cook
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH] scsi: qla2xxx: Fix flex array member not at end
2026-07-15 16:16 ` Kees Cook
@ 2026-07-17 15:45 ` Jesse Taube
0 siblings, 0 replies; 3+ messages in thread
From: Jesse Taube @ 2026-07-17 15:45 UTC (permalink / raw)
To: Kees Cook
Cc: linux-kernel, linux-scsi, Nilesh Javali,
GR-QLogic-Storage-Upstream, James E.J. Bottomley,
Martin K. Petersen, Gustavo A. R. Silva, Christoph Hellwig,
John Meneghini, Chris Leech, Jesse Taube
On Wed, Jul 15, 2026 at 12:16 PM Kees Cook <kees@kernel.org> wrote:
>
> On Tue, Jul 14, 2026 at 11:22:38AM -0400, Jesse Taube wrote:
> > In qla_edif_bsg.h: `struct fc_bsg_reply` and `struct fc_bsg_request`
> > have flexible array members, thus they must be the last member of
> > the parent structure. Contininging in the effort to add
> > `-Wflex-array-member-not-at-end`, move the structs to the end of
> > the parent structures, `struct qla_bsg_auth_els_reply` and
> > `struct qla_bsg_auth_els_request `.
> >
> > Suggested-by: John Meneghini <jmeneghi@redhat.com>
> > Signed-off-by: Jesse Taube <jtaubepe@redhat.com>
>
> It feels like this structure is part of the hardware spec? (and therefore
> cannot be re-ordered.) I'm not sure at all, though, but it looks like
> it's part of request serialization:
I looked at the code closer and i relized this structure is passed to userspace.
I sent a patch V2 after i relized that though i accidentally kept the
commit description the same.
In the lpfc driver here you can see how `bsg_job->reply` is supossed to be used.
https://elixir.bootlin.com/linux/v7.0.10/source/drivers/scsi/lpfc/lpfc_bsg.c#L662
the layout of `bsg_job->reply` is defined in `struct fc_bsg_reply`,
the `struct qla_bsg_auth_els_request`
data should have been put in the flex-array called `vendor_rsp` so
`(struct qla_bsg_auth_els_request *)bsg_job->request->vendor_reply->vendor_rsp;
unfortunatly because of `vendor_reply` and `ctels_reply` being in a
union the actuall vendor data from
`struct qla_bsg_auth_els_request` is offset by the size of
`ctels_reply`, this can be seen in V2 of this patch
AFAIK this isn't leaking data.
V2: https://lore.kernel.org/all/20260714185621.610105-1-jtaubepe@redhat.com/
Thanks,
Jesse Taube
> $ git grep 'qla_bsg_auth_els_request' | grep '>request'
> drivers/scsi/qla2xxx/qla_edif.c: (struct qla_bsg_auth_els_request *)bsg_job->request;
> drivers/scsi/qla2xxx/qla_edif.c: (struct qla_bsg_auth_els_request *)bsg_job->request;
> drivers/scsi/qla2xxx/qla_edif.c: (struct qla_bsg_auth_els_request *)bsg_job->request;
> drivers/scsi/qla2xxx/qla_isr.c: (struct qla_bsg_auth_els_request *)bsg_job->request;
>
> -Kees
>
> > ---
> > drivers/scsi/qla2xxx/qla_edif_bsg.h | 4 ++--
> > 1 file changed, 2 insertions(+), 2 deletions(-)
> >
> > diff --git a/drivers/scsi/qla2xxx/qla_edif_bsg.h b/drivers/scsi/qla2xxx/qla_edif_bsg.h
> > index 514c265ba86e..bcc35d149c1f 100644
> > --- a/drivers/scsi/qla2xxx/qla_edif_bsg.h
> > +++ b/drivers/scsi/qla2xxx/qla_edif_bsg.h
> > @@ -36,16 +36,16 @@ struct extra_auth_els {
> > } __packed;
> >
> > struct qla_bsg_auth_els_request {
> > - struct fc_bsg_request r;
> > struct extra_auth_els e;
> > + struct fc_bsg_request r;
> > };
> >
> > struct qla_bsg_auth_els_reply {
> > - struct fc_bsg_reply r;
> > uint32_t rx_xchg_address;
> > uint8_t version;
> > uint8_t pad[VND_CMD_PAD_SIZE];
> > uint8_t reserved[VND_CMD_APP_RESERVED_SIZE];
> > + struct fc_bsg_reply r;
> > };
> >
> > struct app_id {
> > --
> > 2.54.0
> >
>
> --
> Kees Cook
>
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2026-07-17 15:45 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2026-07-14 15:22 [PATCH] scsi: qla2xxx: Fix flex array member not at end Jesse Taube
2026-07-15 16:16 ` Kees Cook
2026-07-17 15:45 ` Jesse Taube
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox