diff mbox series

[v2] scsi: fcoe: use sysfs_match_string over fcoe_parse_mode

Message ID 20231212-strncpy-drivers-scsi-fcoe-fcoe_sysfs-c-v2-1-1f2d6b2fc409@google.com (mailing list archive)
State Mainlined
Commit edc22a7c86888d1f2442e359c3b33a861045e9cd
Headers show
Series [v2] scsi: fcoe: use sysfs_match_string over fcoe_parse_mode | expand

Commit Message

Justin Stitt Dec. 12, 2023, 11:19 p.m. UTC 
Instead of copying @buf into a new buffer and carefully managing its
newline/null-terminating status, we can just use sysfs_match_string()
as it uses sysfs_streq() internally which handles newline/null-term:

|  /**
|   * sysfs_streq - return true if strings are equal, modulo trailing newline
|   * @s1: one string
|   * @s2: another string
|   *
|   * This routine returns true iff two strings are equal, treating both
|   * NUL and newline-then-NUL as equivalent string terminations.  It's
|   * geared for use with sysfs input strings, which generally terminate
|   * with newlines but are compared against values without newlines.
|   */
|  bool sysfs_streq(const char *s1, const char *s2)
|  ...

Then entirely drop the now unused fcoe_parse_mode, being careful to
change if condition from checking for FIP_CONN_TYPE_UNKNOWN to < 0 as
sysfs_match_string can return -EINVAL. Also check explicitly if
ctlr->mode is equal to FIP_CONN_TYPE_UNKNOWN -- this is probably
preferred to "<=" as the behavior is more obvious while maintaining
functionality.

To get the compiler not to complain, make fip_conn_type_names
const char * const. Perhaps, this should also be done for
fcf_state_names.

This also removes an instance of strncpy() which helps [1].

Link: https://github.com/KSPP/linux/issues/90 [1]
Cc: linux-hardening@vger.kernel.org
Signed-off-by: Justin Stitt <justinstitt@google.com>
---
Changes in v2:
- update if-cond to check for unknown type (thanks Kees)
- Link to v1: https://lore.kernel.org/r/20231211-strncpy-drivers-scsi-fcoe-fcoe_sysfs-c-v1-1-73b942238396@google.com
---
Builds upon patch and feedback from [2]:

However, this is different enough to warrant its own patch and not be a
continuation.

[2]: https://lore.kernel.org/all/9f38f4aa-c6b5-4786-a641-d02d8bd92f7f@acm.org/
---
 drivers/scsi/fcoe/fcoe_sysfs.c | 26 ++++----------------------
 1 file changed, 4 insertions(+), 22 deletions(-)


---
base-commit: bee0e7762ad2c6025b9f5245c040fcc36ef2bde8
change-id: 20231024-strncpy-drivers-scsi-fcoe-fcoe_sysfs-c-0e1dffe82855

Best regards,
--
Justin Stitt <justinstitt@google.com>

Comments

Kees Cook Dec. 13, 2023, 2:33 a.m. UTC | #1 
On Tue, Dec 12, 2023 at 11:19:06PM +0000, Justin Stitt wrote:
> Instead of copying @buf into a new buffer and carefully managing its
> newline/null-terminating status, we can just use sysfs_match_string()
> as it uses sysfs_streq() internally which handles newline/null-term:
> 
> |  /**
> |   * sysfs_streq - return true if strings are equal, modulo trailing newline
> |   * @s1: one string
> |   * @s2: another string
> |   *
> |   * This routine returns true iff two strings are equal, treating both
> |   * NUL and newline-then-NUL as equivalent string terminations.  It's
> |   * geared for use with sysfs input strings, which generally terminate
> |   * with newlines but are compared against values without newlines.
> |   */
> |  bool sysfs_streq(const char *s1, const char *s2)
> |  ...
> 
> Then entirely drop the now unused fcoe_parse_mode, being careful to
> change if condition from checking for FIP_CONN_TYPE_UNKNOWN to < 0 as
> sysfs_match_string can return -EINVAL. Also check explicitly if
> ctlr->mode is equal to FIP_CONN_TYPE_UNKNOWN -- this is probably
> preferred to "<=" as the behavior is more obvious while maintaining
> functionality.
> 
> To get the compiler not to complain, make fip_conn_type_names
> const char * const. Perhaps, this should also be done for
> fcf_state_names.
> 
> This also removes an instance of strncpy() which helps [1].
> 
> Link: https://github.com/KSPP/linux/issues/90 [1]
> Cc: linux-hardening@vger.kernel.org
> Signed-off-by: Justin Stitt <justinstitt@google.com>

Looks great; thanks!

Reviewed-by: Kees Cook <keescook@chromium.org>
Hannes Reinecke Dec. 13, 2023, 7:07 a.m. UTC | #2 
On 12/13/23 00:19, Justin Stitt wrote:
> Instead of copying @buf into a new buffer and carefully managing its
> newline/null-terminating status, we can just use sysfs_match_string()
> as it uses sysfs_streq() internally which handles newline/null-term:
> 
> |  /**
> |   * sysfs_streq - return true if strings are equal, modulo trailing newline
> |   * @s1: one string
> |   * @s2: another string
> |   *
> |   * This routine returns true iff two strings are equal, treating both
> |   * NUL and newline-then-NUL as equivalent string terminations.  It's
> |   * geared for use with sysfs input strings, which generally terminate
> |   * with newlines but are compared against values without newlines.
> |   */
> |  bool sysfs_streq(const char *s1, const char *s2)
> |  ...
> 
> Then entirely drop the now unused fcoe_parse_mode, being careful to
> change if condition from checking for FIP_CONN_TYPE_UNKNOWN to < 0 as
> sysfs_match_string can return -EINVAL. Also check explicitly if
> ctlr->mode is equal to FIP_CONN_TYPE_UNKNOWN -- this is probably
> preferred to "<=" as the behavior is more obvious while maintaining
> functionality.
> 
> To get the compiler not to complain, make fip_conn_type_names
> const char * const. Perhaps, this should also be done for
> fcf_state_names.
> 
> This also removes an instance of strncpy() which helps [1].
> 
> Link: https://github.com/KSPP/linux/issues/90 [1]
> Cc: linux-hardening@vger.kernel.org
> Signed-off-by: Justin Stitt <justinstitt@google.com>
> ---
> Changes in v2:
> - update if-cond to check for unknown type (thanks Kees)
> - Link to v1: https://lore.kernel.org/r/20231211-strncpy-drivers-scsi-fcoe-fcoe_sysfs-c-v1-1-73b942238396@google.com
> ---
> Builds upon patch and feedback from [2]:
> 
> However, this is different enough to warrant its own patch and not be a
> continuation.
> 
> [2]: https://lore.kernel.org/all/9f38f4aa-c6b5-4786-a641-d02d8bd92f7f@acm.org/
> ---
>   drivers/scsi/fcoe/fcoe_sysfs.c | 26 ++++----------------------
>   1 file changed, 4 insertions(+), 22 deletions(-)
> 
Reviewed-by: Hannes Reinecke <hare@suse.de>

Cheers,

Hannes
Martin K. Petersen Dec. 14, 2023, 3:25 a.m. UTC | #3 
Justin,

> Instead of copying @buf into a new buffer and carefully managing its
> newline/null-terminating status, we can just use sysfs_match_string()
> as it uses sysfs_streq() internally which handles newline/null-term:

Applied to 6.8/scsi-staging, thanks!
Martin K. Petersen Dec. 19, 2023, 2:18 a.m. UTC | #4 
On Tue, 12 Dec 2023 23:19:06 +0000, Justin Stitt wrote:

> Instead of copying @buf into a new buffer and carefully managing its
> newline/null-terminating status, we can just use sysfs_match_string()
> as it uses sysfs_streq() internally which handles newline/null-term:
> 
> |  /**
> |   * sysfs_streq - return true if strings are equal, modulo trailing newline
> |   * @s1: one string
> |   * @s2: another string
> |   *
> |   * This routine returns true iff two strings are equal, treating both
> |   * NUL and newline-then-NUL as equivalent string terminations.  It's
> |   * geared for use with sysfs input strings, which generally terminate
> |   * with newlines but are compared against values without newlines.
> |   */
> |  bool sysfs_streq(const char *s1, const char *s2)
> |  ...
> 
> [...]

Applied to 6.8/scsi-queue, thanks!

[1/1] scsi: fcoe: use sysfs_match_string over fcoe_parse_mode
      https://git.kernel.org/mkp/scsi/c/edc22a7c8688
diff mbox series

Patch

diff --git a/drivers/scsi/fcoe/fcoe_sysfs.c b/drivers/scsi/fcoe/fcoe_sysfs.c
index e17957f8085c..408a806bf4c2 100644
--- a/drivers/scsi/fcoe/fcoe_sysfs.c
+++ b/drivers/scsi/fcoe/fcoe_sysfs.c
@@ -10,6 +10,7 @@ 
 #include <linux/kernel.h>
 #include <linux/etherdevice.h>
 #include <linux/ctype.h>
+#include <linux/string.h>
 
 #include <scsi/fcoe_sysfs.h>
 #include <scsi/libfcoe.h>
@@ -214,25 +215,13 @@  static const char *get_fcoe_##title##_name(enum table_type table_key)	\
 	return table[table_key];					\
 }
 
-static char *fip_conn_type_names[] = {
+static const char * const fip_conn_type_names[] = {
 	[ FIP_CONN_TYPE_UNKNOWN ] = "Unknown",
 	[ FIP_CONN_TYPE_FABRIC ]  = "Fabric",
 	[ FIP_CONN_TYPE_VN2VN ]   = "VN2VN",
 };
 fcoe_enum_name_search(ctlr_mode, fip_conn_type, fip_conn_type_names)
 
-static enum fip_conn_type fcoe_parse_mode(const char *buf)
-{
-	int i;
-
-	for (i = 0; i < ARRAY_SIZE(fip_conn_type_names); i++) {
-		if (strcasecmp(buf, fip_conn_type_names[i]) == 0)
-			return i;
-	}
-
-	return FIP_CONN_TYPE_UNKNOWN;
-}
-
 static char *fcf_state_names[] = {
 	[ FCOE_FCF_STATE_UNKNOWN ]      = "Unknown",
 	[ FCOE_FCF_STATE_DISCONNECTED ] = "Disconnected",
@@ -274,17 +263,10 @@  static ssize_t store_ctlr_mode(struct device *dev,
 			       const char *buf, size_t count)
 {
 	struct fcoe_ctlr_device *ctlr = dev_to_ctlr(dev);
-	char mode[FCOE_MAX_MODENAME_LEN + 1];
 
 	if (count > FCOE_MAX_MODENAME_LEN)
 		return -EINVAL;
 
-	strncpy(mode, buf, count);
-
-	if (mode[count - 1] == '\n')
-		mode[count - 1] = '\0';
-	else
-		mode[count] = '\0';
 
 	switch (ctlr->enabled) {
 	case FCOE_CTLR_ENABLED:
@@ -297,8 +279,8 @@  static ssize_t store_ctlr_mode(struct device *dev,
 			return -ENOTSUPP;
 		}
 
-		ctlr->mode = fcoe_parse_mode(mode);
-		if (ctlr->mode == FIP_CONN_TYPE_UNKNOWN) {
+		ctlr->mode = sysfs_match_string(fip_conn_type_names, buf);
+		if (ctlr->mode < 0 || ctlr->mode == FIP_CONN_TYPE_UNKNOWN) {
 			LIBFCOE_SYSFS_DBG(ctlr, "Unknown mode %s provided.\n",
 					  buf);
 			return -EINVAL;