strlcpy() reads the entire source buffer first. This read may exceed
the destination size limit. This is both inefficient and can lead
to linear read overflows if a source string is not NUL-terminated[1].
Additionally, it returns the size of the source string, not the
resulting size of the destination string. In an effort to remove strlcpy()
completely[2], replace strlcpy() here with strscpy().
Be explicitly robust in the face of truncation, which should be an
impossible state.
Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#strlcpy [1]
Link: https://github.com/KSPP/linux/issues/89 [2]
Cc: Steffen Maier <[email protected]>
Cc: Benjamin Block <[email protected]>
Cc: Heiko Carstens <[email protected]>
Cc: Vasily Gorbik <[email protected]>
Cc: Alexander Gordeev <[email protected]>
Cc: Christian Borntraeger <[email protected]>
Cc: Sven Schnelle <[email protected]>
Cc: Azeem Shaikh <[email protected]>
Cc: [email protected]
Signed-off-by: Kees Cook <[email protected]>
---
drivers/s390/scsi/zfcp_fc.c | 11 +++++++++--
1 file changed, 9 insertions(+), 2 deletions(-)
diff --git a/drivers/s390/scsi/zfcp_fc.c b/drivers/s390/scsi/zfcp_fc.c
index 4f0d0e55f0d4..1a29f10767fc 100644
--- a/drivers/s390/scsi/zfcp_fc.c
+++ b/drivers/s390/scsi/zfcp_fc.c
@@ -900,8 +900,15 @@ static void zfcp_fc_rspn(struct zfcp_adapter *adapter,
zfcp_fc_ct_ns_init(&rspn_req->ct_hdr, FC_NS_RSPN_ID,
FC_SYMBOLIC_NAME_SIZE);
hton24(rspn_req->rspn.fr_fid.fp_fid, fc_host_port_id(shost));
- len = strlcpy(rspn_req->rspn.fr_name, fc_host_symbolic_name(shost),
- FC_SYMBOLIC_NAME_SIZE);
+ len = strscpy(rspn_req->name, fc_host_symbolic_name(shost),
+ sizeof(rspn_req->name));
+ /*
+ * It should be impossible for this to truncate, as
+ * sizeof(rspn_req->name) is equal to max size of
+ * fc_host_symbolic_name(shost), but check anyway.
+ */
+ if (len < 0)
+ len = sizeof(rspn_req->name) - 1;
rspn_req->rspn.fr_name_len = len;
sg_init_one(&fc_req->sg_req, rspn_req, sizeof(*rspn_req));
--
2.34.1
On Thu, Nov 16, 2023 at 11:14:35AM -0800, Kees Cook wrote:
Hi Kees,
> diff --git a/drivers/s390/scsi/zfcp_fc.c b/drivers/s390/scsi/zfcp_fc.c
> index 4f0d0e55f0d4..1a29f10767fc 100644
> --- a/drivers/s390/scsi/zfcp_fc.c
> +++ b/drivers/s390/scsi/zfcp_fc.c
> @@ -900,8 +900,15 @@ static void zfcp_fc_rspn(struct zfcp_adapter *adapter,
> zfcp_fc_ct_ns_init(&rspn_req->ct_hdr, FC_NS_RSPN_ID,
> FC_SYMBOLIC_NAME_SIZE);
> hton24(rspn_req->rspn.fr_fid.fp_fid, fc_host_port_id(shost));
> - len = strlcpy(rspn_req->rspn.fr_name, fc_host_symbolic_name(shost),
> - FC_SYMBOLIC_NAME_SIZE);
> + len = strscpy(rspn_req->name, fc_host_symbolic_name(shost),
> + sizeof(rspn_req->name));
Could you please explain why do you copy to rspn_req->name instead
of rspn_req->rspn.fr_name?
> + /*
> + * It should be impossible for this to truncate, as
> + * sizeof(rspn_req->name) is equal to max size of
> + * fc_host_symbolic_name(shost), but check anyway.
> + */
> + if (len < 0)
> + len = sizeof(rspn_req->name) - 1;
> rspn_req->rspn.fr_name_len = len;
>
> sg_init_one(&fc_req->sg_req, rspn_req, sizeof(*rspn_req));
Thanks!
On Fri, Nov 17, 2023 at 07:19:48PM +0100, Alexander Gordeev wrote:
> > @@ -900,8 +900,15 @@ static void zfcp_fc_rspn(struct zfcp_adapter *adapter,
> > zfcp_fc_ct_ns_init(&rspn_req->ct_hdr, FC_NS_RSPN_ID,
> > FC_SYMBOLIC_NAME_SIZE);
> > hton24(rspn_req->rspn.fr_fid.fp_fid, fc_host_port_id(shost));
> > - len = strlcpy(rspn_req->rspn.fr_name, fc_host_symbolic_name(shost),
> > - FC_SYMBOLIC_NAME_SIZE);
> > + len = strscpy(rspn_req->name, fc_host_symbolic_name(shost),
> > + sizeof(rspn_req->name));
>
> Could you please explain why do you copy to rspn_req->name instead
> of rspn_req->rspn.fr_name?
Please, ignore this (stupid Friday evening) quesiton.
Although the use of sizeof() is right thing, FC_SYMBOLIC_NAME_SIZE
is so ubiquotous in this source that it probably makes sense to
address in a separate cleanup.
@Steffen, @Benjamin, could you please comment in this and below?
> > + /*
> > + * It should be impossible for this to truncate, as
> > + * sizeof(rspn_req->name) is equal to max size of
> > + * fc_host_symbolic_name(shost), but check anyway.
> > + */
> > + if (len < 0)
> > + len = sizeof(rspn_req->name) - 1;
Not sure if this check is really needed. It could have been done
for strlcpy() also, but as you say - should not ever happen.
> > rspn_req->rspn.fr_name_len = len;
> >
> > sg_init_one(&fc_req->sg_req, rspn_req, sizeof(*rspn_req));
Thanks!
Hey Kees,
thanks for the patch.
can you please send this patch to linux-scsi and CC the SCSI Maintainers
(Martin and James) instead (having linux-s390 on CC is fine)? zFCP doesn't go
via s390, being a SCSI driver.
On Thu, Nov 16, 2023 at 11:14:35AM -0800, Kees Cook wrote:
> strlcpy() reads the entire source buffer first. This read may exceed
> the destination size limit. This is both inefficient and can lead
> to linear read overflows if a source string is not NUL-terminated[1].
> Additionally, it returns the size of the source string, not the
> resulting size of the destination string. In an effort to remove strlcpy()
> completely[2], replace strlcpy() here with strscpy().
>
> Be explicitly robust in the face of truncation, which should be an
> impossible state.
>
> Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#strlcpy [1]
> Link: https://github.com/KSPP/linux/issues/89 [2]
> Cc: Steffen Maier <[email protected]>
> Cc: Benjamin Block <[email protected]>
> Cc: Heiko Carstens <[email protected]>
> Cc: Vasily Gorbik <[email protected]>
> Cc: Alexander Gordeev <[email protected]>
> Cc: Christian Borntraeger <[email protected]>
> Cc: Sven Schnelle <[email protected]>
> Cc: Azeem Shaikh <[email protected]>
> Cc: [email protected]
> Signed-off-by: Kees Cook <[email protected]>
> ---
> drivers/s390/scsi/zfcp_fc.c | 11 +++++++++--
> 1 file changed, 9 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/s390/scsi/zfcp_fc.c b/drivers/s390/scsi/zfcp_fc.c
> index 4f0d0e55f0d4..1a29f10767fc 100644
> --- a/drivers/s390/scsi/zfcp_fc.c
> +++ b/drivers/s390/scsi/zfcp_fc.c
> @@ -900,8 +900,15 @@ static void zfcp_fc_rspn(struct zfcp_adapter *adapter,
> zfcp_fc_ct_ns_init(&rspn_req->ct_hdr, FC_NS_RSPN_ID,
> FC_SYMBOLIC_NAME_SIZE);
> hton24(rspn_req->rspn.fr_fid.fp_fid, fc_host_port_id(shost));
> - len = strlcpy(rspn_req->rspn.fr_name, fc_host_symbolic_name(shost),
> - FC_SYMBOLIC_NAME_SIZE);
> + len = strscpy(rspn_req->name, fc_host_symbolic_name(shost),
> + sizeof(rspn_req->name));
This is corrct.
> + /*
> + * It should be impossible for this to truncate, as
> + * sizeof(rspn_req->name) is equal to max size of
> + * fc_host_symbolic_name(shost), but check anyway.
> + */
> + if (len < 0)
> + len = sizeof(rspn_req->name) - 1;
I'd rather have a compile-time check, whether the array sizes of
`rspn_req->name` and `fc_host_symbolic_name(shost)` run out of sync, or
against our expectations.
Something like:
len = strscpy(rspn_req->name, fc_host_symbolic_name(shost),
sizeof(rspn_req->name));
BUILD_BUG_ON(sizeof(rspn_req->name) !=
sizeof(fc_host_symbolic_name(shost)) ||
sizeof(rspn_req->name) !=
1 << sizeof(rspn_req->rspn.fr_name_len) * 8);
rspn_req->rspn.fr_name_len = len;
Then the last assignment should also be correct; and if something changes -
unexpectedly, because this follows a FC standard - we need to adapt the code
anyway I'd think.
Or am I missing something?
> rspn_req->rspn.fr_name_len = len;
>
> sg_init_one(&fc_req->sg_req, rspn_req, sizeof(*rspn_req));
> --
> 2.34.1
>
--
Best Regards, Benjamin Block / Linux on IBM Z Kernel Development
IBM Deutschland Research & Development GmbH / https://www.ibm.com/privacy
Vors. Aufs.-R.: Gregor Pillen / Gesch?ftsf?hrung: David Faller
Sitz der Ges.: B?blingen / Registergericht: AmtsG Stuttgart, HRB 243294