trivial sparse fixes for rdma-core

trivial sparse fixes for rdma-core

[Christoph Hellwig]

This is the low hanging fruit in terms of reducing sparse noise before
looking into endianess and iomem annotations.

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 01/32] libibumad: include umad_str.h in umad_str.c

[Christoph Hellwig]

So that we have the prototypes available in the implementation file
and do get proper type checking for them.

Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
---
 libibumad/umad_str.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/libibumad/umad_str.c b/libibumad/umad_str.c
index 79d51af..d4832e1 100644
--- a/libibumad/umad_str.c
+++ b/libibumad/umad_str.c
@@ -40,6 +40,7 @@
 #include <infiniband/umad_sm.h>
 #include <infiniband/umad_sa.h>
 #include <infiniband/umad_cm.h>
+#include "umad_str.h"
 
 const char * umad_class_str(uint8_t mgmt_class)
 {
-- 
2.1.4

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH 01/32] libibumad: include umad_str.h in umad_str.c

[Hal Rosenstock]

On 10/13/2016 1:51 PM, Christoph Hellwig wrote:
> So that we have the prototypes available in the implementation file
> and do get proper type checking for them.
> 
> Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>

Reviewed-by: Hal Rosenstock <hal-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 02/32] libibumad: create sysfs.h instead of using extern declarations in umad.c

[Christoph Hellwig]

So that we have the prototypes available in the implementation file
and do get proper type checking for them.

Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
---
 libibumad/sysfs.c |  1 +
 libibumad/sysfs.h | 10 ++++++++++
 libibumad/umad.c  |  7 +------
 3 files changed, 12 insertions(+), 6 deletions(-)
 create mode 100644 libibumad/sysfs.h

diff --git a/libibumad/sysfs.c b/libibumad/sysfs.c
index 011e411..83f89b4 100644
--- a/libibumad/sysfs.c
+++ b/libibumad/sysfs.c
@@ -44,6 +44,7 @@
 #include <byteswap.h>
 #include <netinet/in.h>
 #include <infiniband/arch.h>
+#include "sysfs.h"
 
 static int ret_code(void)
 {
diff --git a/libibumad/sysfs.h b/libibumad/sysfs.h
new file mode 100644
index 0000000..7f8cbfe
--- /dev/null
+++ b/libibumad/sysfs.h
@@ -0,0 +1,10 @@
+#ifndef _UMAD_SYSFS_H
+#define _UMAD_SYSFS_H
+
+extern int sys_read_string(const char *dir_name, const char *file_name, char *str, int len);
+extern int sys_read_guid(const char *dir_name, const char *file_name, uint64_t * net_guid);
+extern int sys_read_gid(const char *dir_name, const char *file_name, uint8_t * gid);
+extern int sys_read_uint64(const char *dir_name, const char *file_name, uint64_t * u);
+extern int sys_read_uint(const char *dir_name, const char *file_name, unsigned *u);
+
+#endif /* _UMAD_SYSFS_H */
diff --git a/libibumad/umad.c b/libibumad/umad.c
index 0c969f1..8860b99 100644
--- a/libibumad/umad.c
+++ b/libibumad/umad.c
@@ -53,6 +53,7 @@
 #define IB_OPENIB_OUI                 (0x001405)
 
 #include <valgrind/memcheck.h>
+#include "sysfs.h"
 
 typedef struct ib_user_mad_reg_req {
 	uint32_t id;
@@ -77,12 +78,6 @@ struct ib_user_mad_reg_req2 {
 	uint8_t  reserved[3];
 };
 
-extern int sys_read_string(const char *dir_name, const char *file_name, char *str, int len);
-extern int sys_read_guid(const char *dir_name, const char *file_name, uint64_t * net_guid);
-extern int sys_read_gid(const char *dir_name, const char *file_name, uint8_t * gid);
-extern int sys_read_uint64(const char *dir_name, const char *file_name, uint64_t * u);
-extern int sys_read_uint(const char *dir_name, const char *file_name, unsigned *u);
-
 #define IBWARN(fmt, args...) fprintf(stderr, "ibwarn: [%d] %s: " fmt "\n", getpid(), __func__, ## args)
 
 #define TRACE	if (umaddebug)	IBWARN
-- 
2.1.4

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH 02/32] libibumad: create sysfs.h instead of using extern declarations in umad.c

[Hal Rosenstock]

On 10/13/2016 1:51 PM, Christoph Hellwig wrote:
> So that we have the prototypes available in the implementation file
> and do get proper type checking for them.
> 
> Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>

Reviewed-by: Hal Rosenstock <hal-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>

> ---
>  libibumad/sysfs.c |  1 +
>  libibumad/sysfs.h | 10 ++++++++++
>  libibumad/umad.c  |  7 +------
>  3 files changed, 12 insertions(+), 6 deletions(-)
>  create mode 100644 libibumad/sysfs.h
> 
> diff --git a/libibumad/sysfs.c b/libibumad/sysfs.c
> index 011e411..83f89b4 100644
> --- a/libibumad/sysfs.c
> +++ b/libibumad/sysfs.c
> @@ -44,6 +44,7 @@
>  #include <byteswap.h>
>  #include <netinet/in.h>
>  #include <infiniband/arch.h>
> +#include "sysfs.h"
>  
>  static int ret_code(void)
>  {
> diff --git a/libibumad/sysfs.h b/libibumad/sysfs.h
> new file mode 100644
> index 0000000..7f8cbfe
> --- /dev/null
> +++ b/libibumad/sysfs.h
> @@ -0,0 +1,10 @@
> +#ifndef _UMAD_SYSFS_H
> +#define _UMAD_SYSFS_H

What about copyright for this new file ?

> +
> +extern int sys_read_string(const char *dir_name, const char *file_name, char *str, int len);
> +extern int sys_read_guid(const char *dir_name, const char *file_name, uint64_t * net_guid);
> +extern int sys_read_gid(const char *dir_name, const char *file_name, uint8_t * gid);
> +extern int sys_read_uint64(const char *dir_name, const char *file_name, uint64_t * u);
> +extern int sys_read_uint(const char *dir_name, const char *file_name, unsigned *u);
> +
> +#endif /* _UMAD_SYSFS_H */
> diff --git a/libibumad/umad.c b/libibumad/umad.c
> index 0c969f1..8860b99 100644
> --- a/libibumad/umad.c
> +++ b/libibumad/umad.c
> @@ -53,6 +53,7 @@
>  #define IB_OPENIB_OUI                 (0x001405)
>  
>  #include <valgrind/memcheck.h>
> +#include "sysfs.h"
>  
>  typedef struct ib_user_mad_reg_req {
>  	uint32_t id;
> @@ -77,12 +78,6 @@ struct ib_user_mad_reg_req2 {
>  	uint8_t  reserved[3];
>  };
>  
> -extern int sys_read_string(const char *dir_name, const char *file_name, char *str, int len);
> -extern int sys_read_guid(const char *dir_name, const char *file_name, uint64_t * net_guid);
> -extern int sys_read_gid(const char *dir_name, const char *file_name, uint8_t * gid);
> -extern int sys_read_uint64(const char *dir_name, const char *file_name, uint64_t * u);
> -extern int sys_read_uint(const char *dir_name, const char *file_name, unsigned *u);
> -
>  #define IBWARN(fmt, args...) fprintf(stderr, "ibwarn: [%d] %s: " fmt "\n", getpid(), __func__, ## args)
>  
>  #define TRACE	if (umaddebug)	IBWARN
> 
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 03/32] libibumad: use NULL instead of 0 for NULL pointers

[Christoph Hellwig]

Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
---
 libibumad/sysfs.c |  8 ++++----
 libibumad/umad.c  | 17 +++++++++--------
 2 files changed, 13 insertions(+), 12 deletions(-)

diff --git a/libibumad/sysfs.c b/libibumad/sysfs.c
index 83f89b4..730374d 100644
--- a/libibumad/sysfs.c
+++ b/libibumad/sysfs.c
@@ -95,7 +95,7 @@ int sys_read_guid(const char *dir_name, const char *file_name, uint64_t * net_gu
 	for (s = buf, i = 0; i < 4; i++) {
 		if (!(str = strsep(&s, ": \t\n")))
 			return -EINVAL;
-		guid = (guid << 16) | (strtoul(str, 0, 16) & 0xffff);
+		guid = (guid << 16) | (strtoul(str, NULL, 16) & 0xffff);
 	}
 
 	*net_guid = htonll(guid);
@@ -115,7 +115,7 @@ int sys_read_gid(const char *dir_name, const char *file_name, uint8_t * gid)
 	for (s = buf, i = 0; i < 8; i++) {
 		if (!(str = strsep(&s, ": \t\n")))
 			return -EINVAL;
-		ugid[i] = htons(strtoul(str, 0, 16) & 0xffff);
+		ugid[i] = htons(strtoul(str, NULL, 16) & 0xffff);
 	}
 
 	return 0;
@@ -129,7 +129,7 @@ int sys_read_uint64(const char *dir_name, const char *file_name, uint64_t * u)
 	if ((r = sys_read_string(dir_name, file_name, buf, sizeof(buf))) < 0)
 		return r;
 
-	*u = strtoull(buf, 0, 0);
+	*u = strtoull(buf, NULL, 0);
 
 	return 0;
 }
@@ -142,7 +142,7 @@ int sys_read_uint(const char *dir_name, const char *file_name, unsigned *u)
 	if ((r = sys_read_string(dir_name, file_name, buf, sizeof(buf))) < 0)
 		return r;
 
-	*u = strtoul(buf, 0, 0);
+	*u = strtoul(buf, NULL, 0);
 
 	return 0;
 }
diff --git a/libibumad/umad.c b/libibumad/umad.c
index 8860b99..c1b7338 100644
--- a/libibumad/umad.c
+++ b/libibumad/umad.c
@@ -215,7 +215,7 @@ static int release_ca(umad_ca_t * ca)
 			continue;
 		release_port(ca->ports[i]);
 		free(ca->ports[i]);
-		ca->ports[i] = 0;
+		ca->ports[i] = NULL;
 	}
 	return 0;
 }
@@ -320,13 +320,13 @@ static const char *resolve_ca_name(const char *ca_name, int *best_port)
 
 	if (ca_name) {
 		if (resolve_ca_port(ca_name, best_port) < 0)
-			return 0;
+			return NULL;
 		return ca_name;
 	}
 
 	/* Get the list of CA names */
 	if ((n = umad_get_cas_names((void *)names, UMAD_MAX_DEVICES)) < 0)
-		return 0;
+		return NULL;
 
 	/* Find the first existing CA with an active port */
 	for (caidx = 0; caidx < n; caidx++) {
@@ -354,7 +354,8 @@ static const char *resolve_ca_name(const char *ca_name, int *best_port)
 	}
 
 	DEBUG("phys found %d on %s port %d",
-	      phys_found, phys_found >= 0 ? names[phys_found] : 0, port_found);
+	      phys_found, phys_found >= 0 ? names[phys_found] : NULL,
+	      port_found);
 	if (phys_found >= 0) {
 		if (best_port)
 			*best_port = port_found;
@@ -404,7 +405,7 @@ static int get_ca(const char *ca_name, umad_ca_t * ca)
 	if (!(dir = opendir(dir_name)))
 		return -ENOENT;
 
-	if ((r = scandir(dir_name, &namelist, 0, alphasort)) < 0) {
+	if ((r = scandir(dir_name, &namelist, NULL, alphasort)) < 0) {
 		ret = errno < 0 ? errno : -EIO;
 		goto error;
 	}
@@ -544,7 +545,7 @@ int umad_get_cas_names(char cas[][UMAD_CA_NAME_LEN], int max)
 
 	TRACE("max %d", max);
 
-	n = scandir(SYS_INFINIBAND, &namelist, 0, alphasort);
+	n = scandir(SYS_INFINIBAND, &namelist, NULL, alphasort);
 	if (n > 0) {
 		for (i = 0; i < n; i++) {
 			if (strcmp(namelist[i]->d_name, ".") &&
@@ -573,7 +574,7 @@ int umad_get_ca_portguids(const char *ca_name, uint64_t * portguids, int max)
 	int ports = 0, i;
 
 	TRACE("ca name %s max port guids %d", ca_name, max);
-	if (!(ca_name = resolve_ca_name(ca_name, 0)))
+	if (!(ca_name = resolve_ca_name(ca_name, NULL)))
 		return -ENODEV;
 
 	if (umad_get_ca(ca_name, &ca) < 0)
@@ -650,7 +651,7 @@ int umad_get_ca(const char *ca_name, umad_ca_t * ca)
 	int r;
 
 	TRACE("ca_name %s", ca_name);
-	if (!(ca_name = resolve_ca_name(ca_name, 0)))
+	if (!(ca_name = resolve_ca_name(ca_name, NULL)))
 		return -ENODEV;
 
 	if (find_cached_ca(ca_name, ca) > 0)
-- 
2.1.4

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH 03/32] libibumad: use NULL instead of 0 for NULL pointers

[Hal Rosenstock]

On 10/13/2016 1:51 PM, Christoph Hellwig wrote:
> Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>

Reviewed-by: Hal Rosenstock <hal-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 04/32] libibumad: mark symbols as static where possible

[Christoph Hellwig]

Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
---
 libibumad/tests/umad_reg2_compat.c | 10 +++++-----
 libibumad/tests/umad_register2.c   | 12 ++++++------
 libibumad/umad.c                   |  2 +-
 3 files changed, 12 insertions(+), 12 deletions(-)

diff --git a/libibumad/tests/umad_reg2_compat.c b/libibumad/tests/umad_reg2_compat.c
index 264882c..9224372 100644
--- a/libibumad/tests/umad_reg2_compat.c
+++ b/libibumad/tests/umad_reg2_compat.c
@@ -42,7 +42,7 @@
 #define UNLIKELY_MGMT_CLASS 0x2F
 #define UNLIKELY_RMPP_MGMT_CLASS 0x4F
 
-int test_failures = 0;
+static int test_failures = 0;
 
 /** =========================================================================
  * Stolen from OpenSM's register
@@ -69,7 +69,7 @@ static void set_bit64(int b, uint64_t *buf)
 	*addr |= mask;
 }
 
-void dump_reg_attr(struct umad_reg_attr *reg_attr)
+static void dump_reg_attr(struct umad_reg_attr *reg_attr)
 {
 	printf("\nmgmt_class %u\n"
 		"mgmt_class_version %u\n"
@@ -85,7 +85,7 @@ void dump_reg_attr(struct umad_reg_attr *reg_attr)
 	      reg_attr->rmpp_version);
 }
 
-int open_test_device(void)
+static int open_test_device(void)
 {
 	int fd = umad_open_port(NULL, 0);
 	if (fd < 0) {
@@ -96,7 +96,7 @@ int open_test_device(void)
 	return fd;
 }
 
-void test_register(void)
+static void test_register(void)
 {
 	int agent_id;
 	long method_mask[16 / sizeof(long)];
@@ -147,7 +147,7 @@ void test_register(void)
 }
 
 
-void test_fall_back(void)
+static void test_fall_back(void)
 {
 	int rc = 0;
 	struct umad_reg_attr reg_attr;
diff --git a/libibumad/tests/umad_register2.c b/libibumad/tests/umad_register2.c
index 7a9b655..477bd29 100644
--- a/libibumad/tests/umad_register2.c
+++ b/libibumad/tests/umad_register2.c
@@ -57,9 +57,9 @@ struct ib_user_mad_reg_req2 {
 	uint8_t  reserved[3];
 };
 
-int test_failures = 0;
+static int test_failures = 0;
 
-void dump_reg_attr(struct umad_reg_attr *reg_attr)
+static void dump_reg_attr(struct umad_reg_attr *reg_attr)
 {
 	printf("\nmgmt_class %u\n"
 		"mgmt_class_version %u\n"
@@ -75,7 +75,7 @@ void dump_reg_attr(struct umad_reg_attr *reg_attr)
 	      reg_attr->rmpp_version);
 }
 
-int open_test_device(void)
+static int open_test_device(void)
 {
 	int fd = umad_open_port(NULL, 0);
 	if (fd < 0) {
@@ -86,7 +86,7 @@ int open_test_device(void)
 	return fd;
 }
 
-void test_fail(void)
+static void test_fail(void)
 {
 	int rc = 0;
 	struct umad_reg_attr reg_attr;
@@ -194,7 +194,7 @@ out:
 	printf("\n *****\nEnd invalid tests\n");
 }
 
-void test_oui(void)
+static void test_oui(void)
 {
 	int rc = 0;
 	struct umad_reg_attr reg_attr;
@@ -251,7 +251,7 @@ out:
 	printf("\n End valid oui tests\n *****\n");
 }
 
-void check_register2_support(void)
+static void check_register2_support(void)
 {
 	struct ib_user_mad_reg_req2 req;
 	int fd;
diff --git a/libibumad/umad.c b/libibumad/umad.c
index c1b7338..1ac2948 100644
--- a/libibumad/umad.c
+++ b/libibumad/umad.c
@@ -83,7 +83,7 @@ struct ib_user_mad_reg_req2 {
 #define TRACE	if (umaddebug)	IBWARN
 #define DEBUG	if (umaddebug)	IBWARN
 
-int umaddebug = 0;
+static int umaddebug = 0;
 
 #define UMAD_DEV_FILE_SZ	256
 
-- 
2.1.4

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH 04/32] libibumad: mark symbols as static where possible

[Hal Rosenstock]

On 10/13/2016 1:51 PM, Christoph Hellwig wrote:
> Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>

Reviewed-by: Hal Rosenstock <hal-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 05/32] srp_daemon: use NULL instead of 0 for NULL pointers

[Christoph Hellwig]

Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
---
 srp_daemon/srp_daemon.c | 34 +++++++++++++++++-----------------
 1 file changed, 17 insertions(+), 17 deletions(-)

diff --git a/srp_daemon/srp_daemon.c b/srp_daemon/srp_daemon.c
index f16674d..2317508 100644
--- a/srp_daemon/srp_daemon.c
+++ b/srp_daemon/srp_daemon.c
@@ -189,7 +189,7 @@ int srpd_sys_read_gid(char *dir_name, char *file_name, uint8_t *gid)
 	for (s = buf, i = 0 ; i < 8; i++) {
 		if (!(str = strsep(&s, ": \t\n")))
 			return -EINVAL;
-		ugid[i] = htons(strtoul(str, 0, 16) & 0xffff);
+		ugid[i] = htons(strtoul(str, NULL, 16) & 0xffff);
 	}
 
 	return 0;
@@ -203,7 +203,7 @@ int srpd_sys_read_uint64(char *dir_name, char *file_name, uint64_t *u)
 	if ((r = srpd_sys_read_string(dir_name, file_name, buf, sizeof(buf))) < 0)
 		return r;
 
-	*u = strtoull(buf, 0, 0);
+	*u = strtoull(buf, NULL, 0);
 
 	return 0;
 }
@@ -356,36 +356,36 @@ int is_enabled_by_rules_file(struct target_details *target)
 	do {
 		rule++;
 		if (conf->rules[rule].id_ext[0] != '\0' &&
-		    strtoull(target->id_ext, 0, 16) !=
-		    strtoull(conf->rules[rule].id_ext, 0, 16))
+		    strtoull(target->id_ext, NULL, 16) !=
+		    strtoull(conf->rules[rule].id_ext, NULL, 16))
 			continue;
 
 		if (conf->rules[rule].ioc_guid[0] != '\0' &&
 		    ntohll(target->ioc_prof.guid) !=
-		    strtoull(conf->rules[rule].ioc_guid, 0, 16))
+		    strtoull(conf->rules[rule].ioc_guid, NULL, 16))
 			continue;
 
 		if (conf->rules[rule].dgid[0] != '\0') {
 			char tmp = conf->rules[rule].dgid[16];
 			conf->rules[rule].dgid[16] = '\0';
-			if (strtoull(conf->rules[rule].dgid, 0, 16) !=
+			if (strtoull(conf->rules[rule].dgid, NULL, 16) !=
 			    target->subnet_prefix) {
 				conf->rules[rule].dgid[16] = tmp;
 				continue;
 			}
 			conf->rules[rule].dgid[16] = tmp;
-			if (strtoull(&conf->rules[rule].dgid[16], 0, 16) !=
+			if (strtoull(&conf->rules[rule].dgid[16], NULL, 16) !=
 			    target->h_guid)
 				continue;
 		}
 
 		if (conf->rules[rule].service_id[0] != '\0' &&
-		    strtoull(conf->rules[rule].service_id, 0, 16) !=
+		    strtoull(conf->rules[rule].service_id, NULL, 16) !=
 	            target->h_service_id)
 			continue;
 
 		if (conf->rules[rule].pkey[0] != '\0' &&
-		    (uint16_t)strtoul(conf->rules[rule].pkey, 0, 16) !=
+		    (uint16_t)strtoul(conf->rules[rule].pkey, NULL, 16) !=
 	            target->pkey)
 			continue;
 
@@ -436,7 +436,7 @@ static int add_non_exist_target(struct target_details *target)
 		strncpy(subdir_name_ptr, subdir->d_name,
 			MAX_SCSI_HOST_DIR_NAME_LENGTH - prefix_len);
 		if (!check_equal_uint64(scsi_host_dir, "id_ext",
-				        strtoull(target->id_ext, 0, 16)))
+				        strtoull(target->id_ext, NULL, 16)))
 			continue;
 		if (!check_equal_uint16(scsi_host_dir, "pkey", target->pkey) &&
 		    !config->execute)
@@ -1795,7 +1795,7 @@ static int umad_resources_create(struct umad_resources *umad_res)
 
 	umad_res->agent = umad_register(umad_res->portid, SRP_MGMT_CLASS_SA,
 					   SRP_MGMT_CLASS_SA_VERSION,
-					   SRP_SA_RMPP_VERSION, 0);
+					   SRP_SA_RMPP_VERSION, NULL);
 	if (umad_res->agent < 0) {
 		pr_err("umad_register failed\n");
 		return umad_res->agent;
@@ -2064,9 +2064,9 @@ int main(int argc, char *argv[])
 	memset(&sa, 0, sizeof(sa));
 	sigemptyset(&sa.sa_mask);
 	sa.sa_handler = signal_handler;
-	sigaction(SIGINT, &sa, 0);
-	sigaction(SIGTERM, &sa, 0);
-	sigaction(SRP_CATAS_ERR, &sa, 0);
+	sigaction(SIGINT, &sa, NULL);
+	sigaction(SIGTERM, &sa, NULL);
+	sigaction(SRP_CATAS_ERR, &sa, NULL);
 
 	if (strcmp(argv[0] + max_t(int, 0, strlen(argv[0]) - strlen("ibsrpdm")),
 		   "ibsrpdm") == 0) {
@@ -2268,9 +2268,9 @@ close_log:
 	closelog();
 restore_sig:
 	sa.sa_handler = SIG_DFL;
-	sigaction(SIGINT, &sa, 0);
-	sigaction(SIGTERM, &sa, 0);
-	sigaction(SRP_CATAS_ERR, &sa, 0);
+	sigaction(SIGINT, &sa, NULL);
+	sigaction(SIGTERM, &sa, NULL);
+	sigaction(SRP_CATAS_ERR, &sa, NULL);
 close_pipe:
 	close(wakeup_pipe[1]);
 	close(wakeup_pipe[0]);
-- 
2.1.4

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH 05/32] srp_daemon: use NULL instead of 0 for NULL pointers

[Bart Van Assche]

On 10/13/2016 10:51 AM, Christoph Hellwig wrote:
> Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>

Reviewed-by: Bart Van Assche <bart.vanassche-XdAiOPVOjttBDgjK7y7TUQ@public.gmane.org>
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 06/32] srp_daemon: mark symbols as static where possible

[Christoph Hellwig]

Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
---
 srp_daemon/srp_daemon.c | 14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)

diff --git a/srp_daemon/srp_daemon.c b/srp_daemon/srp_daemon.c
index 2317508..f0f05cf 100644
--- a/srp_daemon/srp_daemon.c
+++ b/srp_daemon/srp_daemon.c
@@ -151,7 +151,7 @@ static int check_process_uniqueness(struct config_t *conf)
 	return fd;
 }
 
-int srpd_sys_read_string(const char *dir_name, const char *file_name,
+static int srpd_sys_read_string(const char *dir_name, const char *file_name,
 			 char *str, int max_len)
 {
 	char path[256], *s;
@@ -177,7 +177,7 @@ int srpd_sys_read_string(const char *dir_name, const char *file_name,
 	return 0;
 }
 
-int srpd_sys_read_gid(char *dir_name, char *file_name, uint8_t *gid)
+static int srpd_sys_read_gid(char *dir_name, char *file_name, uint8_t *gid)
 {
 	char buf[64], *str, *s;
 	uint16_t *ugid = (uint16_t *)gid;
@@ -195,7 +195,7 @@ int srpd_sys_read_gid(char *dir_name, char *file_name, uint8_t *gid)
 	return 0;
 }
 
-int srpd_sys_read_uint64(char *dir_name, char *file_name, uint64_t *u)
+static int srpd_sys_read_uint64(char *dir_name, char *file_name, uint64_t *u)
 {
 	char buf[32];
 	int r;
@@ -257,7 +257,7 @@ check_equal_uint16(char *dir_name, char *attr, uint16_t val)
 
 static int recalc(struct resources *res);
 
-void pr_cmd(char *target_str, int not_connected)
+static void pr_cmd(char *target_str, int not_connected)
 {
 	int ret;
 
@@ -343,7 +343,7 @@ static int check_not_equal_int(char *dir_name, char *attr, int value)
 	return 0;
 }
 
-int is_enabled_by_rules_file(struct target_details *target)
+static int is_enabled_by_rules_file(struct target_details *target)
 {
 	int rule;
 	struct config_t *conf = config;
@@ -584,7 +584,7 @@ static int add_non_exist_target(struct target_details *target)
 	return 1;
 }
 
-int send_and_get(int portid, int agent, srp_ib_user_mad_t *out_mad,
+static int send_and_get(int portid, int agent, srp_ib_user_mad_t *out_mad,
 		 srp_ib_user_mad_t *in_mad, int in_mad_size)
 {
 	struct srp_dm_mad *out_dm_mad = (void *) out_mad->hdr.data;
@@ -1804,7 +1804,7 @@ static int umad_resources_create(struct umad_resources *umad_res)
 	return 0;
 }
 
-void *run_thread_retry_to_connect(void *res_in)
+static void *run_thread_retry_to_connect(void *res_in)
 {
 	struct resources *res = (struct resources *)res_in;
 	struct target_details *target;
-- 
2.1.4

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH 06/32] srp_daemon: mark symbols as static where possible

[Bart Van Assche]

On 10/13/2016 10:51 AM, Christoph Hellwig wrote:
> Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>

Reviewed-by: Bart Van Assche <bart.vanassche-XdAiOPVOjttBDgjK7y7TUQ@public.gmane.org>
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 07/32] srp_daemon: give initialize_sysfs a proper ANSI-C prototype

[Christoph Hellwig]

Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
---
 srp_daemon/srp_daemon.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/srp_daemon/srp_daemon.c b/srp_daemon/srp_daemon.c
index f0f05cf..6cb22e4 100644
--- a/srp_daemon/srp_daemon.c
+++ b/srp_daemon/srp_daemon.c
@@ -647,7 +647,7 @@ recv:
 	return -1;
 }
 
-static void initialize_sysfs()
+static void initialize_sysfs(void)
 {
 	char *env;
 
-- 
2.1.4

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH 07/32] srp_daemon: give initialize_sysfs a proper ANSI-C prototype

[Bart Van Assche]

On 10/13/2016 10:51 AM, Christoph Hellwig wrote:
> Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>

This is something I had not yet noticed myself. I think this means we 
will need to enable more compiler warnings for srp_daemon such that gcc 
also catches this. Anyway:

Reviewed-by: Bart Van Assche <bart.vanassche-XdAiOPVOjttBDgjK7y7TUQ@public.gmane.org>
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 08/32] iwpmd: use proper ANSI-C prototypes for functions without arguments

[Christoph Hellwig]

Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
---
 iwpmd/iwarp_pm_common.c |  2 +-
 iwpmd/iwarp_pm_server.c | 10 +++++-----
 2 files changed, 6 insertions(+), 6 deletions(-)

diff --git a/iwpmd/iwarp_pm_common.c b/iwpmd/iwarp_pm_common.c
index 3a885d8..aa6b4a0 100644
--- a/iwpmd/iwarp_pm_common.c
+++ b/iwpmd/iwarp_pm_common.c
@@ -212,7 +212,7 @@ create_socket_v6_exit:
 /**
  * create_netlink_socket - Create netlink socket for the iwarp port mapper
  */
-int create_netlink_socket()
+int create_netlink_socket(void)
 {
 	sockaddr_union bind_addr;
 	struct sockaddr_nl *bind_nl;
diff --git a/iwpmd/iwarp_pm_server.c b/iwpmd/iwarp_pm_server.c
index deed3e7..89d3ee5 100644
--- a/iwpmd/iwarp_pm_server.c
+++ b/iwpmd/iwarp_pm_server.c
@@ -87,7 +87,7 @@ void iwpm_signal_handler(int signum)
 /**
  * iwpm_mapping_reqs_handler - Handle mapping requests timeouts and retries
  */
-void *iwpm_mapping_reqs_handler()
+void *iwpm_mapping_reqs_handler(void)
 {
 	iwpm_mapping_request *iwpm_map_req, *next_map_req;
 	int ret = 0;
@@ -139,7 +139,7 @@ mapping_reqs_handler_exit:
 /**
  * iwpm_pending_msgs_handler - Handle sending iwarp port mapper wire messages
  */
-void *iwpm_pending_msgs_handler()
+void *iwpm_pending_msgs_handler(void)
 {
 	iwpm_pending_msg *pending_msg;
 	iwpm_send_msg *send_msg;
@@ -1279,7 +1279,7 @@ static void iwpm_cleanup(void)
 /**
  * iwarp_port_mapper - Distribute work orders for processing different types of iwpm messages
  */
-static int iwarp_port_mapper()
+static int iwarp_port_mapper(void)
 {
 	fd_set select_fdset; /* read fdset */
 	struct timeval select_timeout;
@@ -1355,8 +1355,8 @@ iwarp_port_mapper_exit:
 
 /**
  * daemonize_iwpm_server - Make iwarp port mapper a daemon process
- */
-static void daemonize_iwpm_server()
+ */ 
+static void daemonize_iwpm_server(void)
 {
 	pid_t pid, sid;
 
-- 
2.1.4

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

RE: [PATCH 08/32] iwpmd: use proper ANSI-C prototypes for functions without arguments

[Steve Wise]

> -----Original Message-----
> From: linux-rdma-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org [mailto:linux-rdma-
> owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org] On Behalf Of Christoph Hellwig
> Sent: Thursday, October 13, 2016 12:51 PM
> To: linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> Subject: [PATCH 08/32] iwpmd: use proper ANSI-C prototypes for functions
without
> arguments
> 
> Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>

Is this fixing some warning?

Looks fine.

Reviewed-by: Steve Wise <swise-7bPotxP6k4+P2YhJcF5u+vpXobYPEAuW@public.gmane.org>



--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH 08/32] iwpmd: use proper ANSI-C prototypes for functions without arguments

['Christoph Hellwig']

On Thu, Oct 13, 2016 at 01:06:25PM -0500, Steve Wise wrote:
> Is this fixing some warning?

Yes, this fixes warnings when running the code through sparse:

../iwpmd/iwarp_pm_common.c:215:27: warning: non-ANSI function declaration of
function 'create_netlink_socket'
../iwpmd/iwarp_pm_server.c:90:33: warning: non-ANSI function declaration of
function 'iwpm_mapping_reqs_handler'
../iwpmd/iwarp_pm_server.c:142:33: warning: non-ANSI function declaration of
function 'iwpm_pending_msgs_handler'
../iwpmd/iwarp_pm_server.c:1282:30: warning: non-ANSI function declaration of
function 'iwarp_port_mapper'
../iwpmd/iwarp_pm_server.c:1359:35: warning: non-ANSI function declaration of
function 'daemonize_iwpm_server'
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH 08/32] iwpmd: use proper ANSI-C prototypes for functions without arguments

[Jason Gunthorpe]

On Thu, Oct 13, 2016 at 08:09:59PM +0200, 'Christoph Hellwig' wrote:
> On Thu, Oct 13, 2016 at 01:06:25PM -0500, Steve Wise wrote:
> > Is this fixing some warning?
> 
> Yes, this fixes warnings when running the code through sparse:
> 
> ../iwpmd/iwarp_pm_common.c:215:27: warning: non-ANSI function declaration of
> function 'create_netlink_socket'

Once these are all fixed, lets turn on -Wstrict-prototypes for gcc?

Jason
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH 08/32] iwpmd: use proper ANSI-C prototypes for functions without arguments

['Christoph Hellwig']

On Thu, Oct 13, 2016 at 12:24:20PM -0600, Jason Gunthorpe wrote:
> On Thu, Oct 13, 2016 at 08:09:59PM +0200, 'Christoph Hellwig' wrote:
> > On Thu, Oct 13, 2016 at 01:06:25PM -0500, Steve Wise wrote:
> > > Is this fixing some warning?
> > 
> > Yes, this fixes warnings when running the code through sparse:
> > 
> > ../iwpmd/iwarp_pm_common.c:215:27: warning: non-ANSI function declaration of
> > function 'create_netlink_socket'
> 
> Once these are all fixed, lets turn on -Wstrict-prototypes for gcc?

Those are fixed with my full series.  Which you won't see for now
as the mail server crapped out after the first 15, so I'll let people
review and merge that batch before spamming the list again.
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 09/32] iwpmd: remove the unused check_iwpm_nlattr_tb prototype

[Christoph Hellwig]

Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
---
 iwpmd/iwarp_pm.h | 2 --
 1 file changed, 2 deletions(-)

diff --git a/iwpmd/iwarp_pm.h b/iwpmd/iwarp_pm.h
index 8b7eeb1..1a2b69a 100644
--- a/iwpmd/iwarp_pm.h
+++ b/iwpmd/iwarp_pm.h
@@ -209,8 +209,6 @@ int create_netlink_socket(void);
 
 void destroy_iwpm_socket(int);
 
-int check_iwpm_nlattr_tb(struct nlattr * [], int);
-
 int parse_iwpm_nlmsg(struct nlmsghdr *, int, struct nla_policy *, struct nlattr * [], const char *);
 
 int parse_iwpm_msg(iwpm_wire_msg *, iwpm_msg_parms *);
-- 
2.1.4

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

RE: [PATCH 09/32] iwpmd: remove the unused check_iwpm_nlattr_tb prototype

[Steve Wise]

> -----Original Message-----
> From: linux-rdma-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org [mailto:linux-rdma-
> owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org] On Behalf Of Christoph Hellwig
> Sent: Thursday, October 13, 2016 12:51 PM
> To: linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> Subject: [PATCH 09/32] iwpmd: remove the unused check_iwpm_nlattr_tb
> prototype
> 
> Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>

Looks good.

Reviewed-by: Steve Wise <swise-7bPotxP6k4+P2YhJcF5u+vpXobYPEAuW@public.gmane.org>


--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 10/32] iwpmd: move external declarations to iwarp_pm.h

[Christoph Hellwig]

Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
---
 iwpmd/iwarp_pm.h        | 8 ++++++++
 iwpmd/iwarp_pm_common.c | 2 --
 iwpmd/iwarp_pm_helper.c | 6 ------
 3 files changed, 8 insertions(+), 8 deletions(-)

diff --git a/iwpmd/iwarp_pm.h b/iwpmd/iwarp_pm.h
index 1a2b69a..766e34f 100644
--- a/iwpmd/iwarp_pm.h
+++ b/iwpmd/iwarp_pm.h
@@ -277,4 +277,12 @@ void free_iwpm_mapped_ports(void);
 extern struct list_head pending_messages;
 extern struct list_head mapping_reqs;
 
+extern iwpm_client client_list[IWARP_PM_MAX_CLIENTS];
+
+extern pthread_cond_t cond_req_complete;
+extern pthread_mutex_t map_req_mutex;
+extern int wake;
+extern pthread_cond_t cond_pending_msg;
+extern pthread_mutex_t pending_msg_mutex;
+
 #endif
diff --git a/iwpmd/iwarp_pm_common.c b/iwpmd/iwarp_pm_common.c
index aa6b4a0..b20a368 100644
--- a/iwpmd/iwarp_pm_common.c
+++ b/iwpmd/iwarp_pm_common.c
@@ -44,8 +44,6 @@ char * iwpm_param_names[IWPM_PARAM_NUM] =
 int iwpm_param_vals[IWPM_PARAM_NUM] = 
 	{ 0 };
 
-extern iwpm_client client_list[IWARP_PM_MAX_CLIENTS];
-
 /**
  * get_iwpm_param()
  */
diff --git a/iwpmd/iwarp_pm_helper.c b/iwpmd/iwarp_pm_helper.c
index 0b115c8..63530f1 100644
--- a/iwpmd/iwarp_pm_helper.c
+++ b/iwpmd/iwarp_pm_helper.c
@@ -35,12 +35,6 @@
 
 static LIST_HEAD(mapped_ports);		/* list of mapped ports */
 
-extern pthread_cond_t cond_req_complete;
-extern pthread_mutex_t map_req_mutex;
-extern int wake; /* set if map_req_thread is wake */
-extern pthread_cond_t cond_pending_msg;
-extern pthread_mutex_t pending_msg_mutex;
-
 /**
  * create_iwpm_map_request - Create a new map request tracking object
  * @req_nlh: netlink header of the received client message
-- 
2.1.4

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

RE: [PATCH 10/32] iwpmd: move external declarations to iwarp_pm.h

[Steve Wise]

> Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>

looks fine.

Reviewed-by: Steve Wise <swise-7bPotxP6k4+P2YhJcF5u+vpXobYPEAuW@public.gmane.org>


--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 11/32] iwpmd: mark symbols static where possible

[Christoph Hellwig]

Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
---
 iwpmd/iwarp_pm_common.c |  6 +++---
 iwpmd/iwarp_pm_server.c | 20 ++++++++++----------
 2 files changed, 13 insertions(+), 13 deletions(-)

diff --git a/iwpmd/iwarp_pm_common.c b/iwpmd/iwarp_pm_common.c
index b20a368..ee9d119 100644
--- a/iwpmd/iwarp_pm_common.c
+++ b/iwpmd/iwarp_pm_common.c
@@ -39,9 +39,9 @@
 #endif
 
 /* iwpm config params */
-char * iwpm_param_names[IWPM_PARAM_NUM] = 
+static char * iwpm_param_names[IWPM_PARAM_NUM] =
 	{ "nl_sock_rbuf_size" };
-int iwpm_param_vals[IWPM_PARAM_NUM] = 
+static int iwpm_param_vals[IWPM_PARAM_NUM] =
 	{ 0 };
 
 /**
@@ -276,7 +276,7 @@ void destroy_iwpm_socket(int pm_sock)
 /**
  * check_iwpm_nlattr - Check for NULL netlink attribute
  */ 
-int check_iwpm_nlattr(struct nlattr *nltb[], int nla_count)
+static int check_iwpm_nlattr(struct nlattr *nltb[], int nla_count)
 { 
         int i, ret = 0;          
         for (i = 1; i < nla_count; i++) {
diff --git a/iwpmd/iwarp_pm_server.c b/iwpmd/iwarp_pm_server.c
index 89d3ee5..b8ae3e3 100644
--- a/iwpmd/iwarp_pm_server.c
+++ b/iwpmd/iwarp_pm_server.c
@@ -34,34 +34,34 @@
 #include "config.h"
 #include "iwarp_pm.h"
 
-const char iwpm_ulib_name [] = "iWarpPortMapperUser";
-int iwpm_version = 3;
+static const char iwpm_ulib_name [] = "iWarpPortMapperUser";
+static int iwpm_version = 3;
 
 LIST_HEAD(mapping_reqs);		      /* list of map tracking objects */
 LIST_HEAD(pending_messages);		      /* list of pending wire messages */
 iwpm_client client_list[IWARP_PM_MAX_CLIENTS];/* list of iwarp port mapper clients */
-int mapinfo_num_list[IWARP_PM_MAX_CLIENTS];   /* list of iwarp port mapper clients */
+static int mapinfo_num_list[IWARP_PM_MAX_CLIENTS];   /* list of iwarp port mapper clients */
 
 /* socket handles */
 static int pmv4_sock, pmv6_sock, netlink_sock, pmv4_client_sock, pmv6_client_sock;
 
-pthread_t map_req_thread; /* handling mapping requests timeout */
-pthread_cond_t cond_req_complete;
+static pthread_t map_req_thread; /* handling mapping requests timeout */
+pthread_cond_t cond_req_complete; 
 pthread_mutex_t map_req_mutex = PTHREAD_MUTEX_INITIALIZER;
 int wake = 0; /* set if map_req_thread is wake */
 
-pthread_t pending_msg_thread; /* sending iwpm wire messages */
+static pthread_t pending_msg_thread; /* sending iwpm wire messages */
 pthread_cond_t cond_pending_msg;
 pthread_mutex_t pending_msg_mutex = PTHREAD_MUTEX_INITIALIZER;
 
 static void iwpm_cleanup(void);
-int print_mappings = 0;
+static int print_mappings = 0;
 
 /**
  * iwpm_signal_handler - Handle signals which iwarp port mapper receives
  * @signum: the number of the caught signal
  */
-void iwpm_signal_handler(int signum)
+static void iwpm_signal_handler(int signum)
 {
 	switch(signum) {
 		case SIGHUP:
@@ -87,7 +87,7 @@ void iwpm_signal_handler(int signum)
 /**
  * iwpm_mapping_reqs_handler - Handle mapping requests timeouts and retries
  */
-void *iwpm_mapping_reqs_handler(void)
+static void *iwpm_mapping_reqs_handler(void)
 {
 	iwpm_mapping_request *iwpm_map_req, *next_map_req;
 	int ret = 0;
@@ -139,7 +139,7 @@ mapping_reqs_handler_exit:
 /**
  * iwpm_pending_msgs_handler - Handle sending iwarp port mapper wire messages
  */
-void *iwpm_pending_msgs_handler(void)
+static void *iwpm_pending_msgs_handler(void)
 {
 	iwpm_pending_msg *pending_msg;
 	iwpm_send_msg *send_msg;
-- 
2.1.4

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

RE: [PATCH 11/32] iwpmd: mark symbols static where possible

[Steve Wise]

> Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>

Looks good

Reviewed-by: Steve Wise <swise-7bPotxP6k4+P2YhJcF5u+vpXobYPEAuW@public.gmane.org>

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 12/32] iwpmd: fix the prototype for the pthread_create start functions

[Christoph Hellwig]

And also remove the unessecary taking of the address for function
arguments while we're at it.

Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
---
 iwpmd/iwarp_pm_server.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/iwpmd/iwarp_pm_server.c b/iwpmd/iwarp_pm_server.c
index b8ae3e3..ab90c6c 100644
--- a/iwpmd/iwarp_pm_server.c
+++ b/iwpmd/iwarp_pm_server.c
@@ -87,7 +87,7 @@ static void iwpm_signal_handler(int signum)
 /**
  * iwpm_mapping_reqs_handler - Handle mapping requests timeouts and retries
  */
-static void *iwpm_mapping_reqs_handler(void)
+static void *iwpm_mapping_reqs_handler(void *unused)
 {
 	iwpm_mapping_request *iwpm_map_req, *next_map_req;
 	int ret = 0;
@@ -139,7 +139,7 @@ mapping_reqs_handler_exit:
 /**
  * iwpm_pending_msgs_handler - Handle sending iwarp port mapper wire messages
  */
-static void *iwpm_pending_msgs_handler(void)
+static void *iwpm_pending_msgs_handler(void *unused)
 {
 	iwpm_pending_msg *pending_msg;
 	iwpm_send_msg *send_msg;
@@ -1438,11 +1438,11 @@ int main(int argc, char *argv[])
 	pthread_cond_init(&cond_req_complete, NULL);
 	pthread_cond_init(&cond_pending_msg, NULL);
 
-	ret = pthread_create(&map_req_thread, NULL, &iwpm_mapping_reqs_handler, NULL);
+	ret = pthread_create(&map_req_thread, NULL, iwpm_mapping_reqs_handler, NULL);
 	if (ret)
 		goto error_exit;
 
-	ret = pthread_create(&pending_msg_thread, NULL, &iwpm_pending_msgs_handler, NULL);
+	ret = pthread_create(&pending_msg_thread, NULL, iwpm_pending_msgs_handler, NULL);
 	if (ret)
 		goto error_exit;
 
-- 
2.1.4

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

RE: [PATCH 12/32] iwpmd: fix the prototype for the pthread_create start functions

[Steve Wise]

> @@ -1438,11 +1438,11 @@ int main(int argc, char *argv[])
>  	pthread_cond_init(&cond_req_complete, NULL);
>  	pthread_cond_init(&cond_pending_msg, NULL);
> 
> -	ret = pthread_create(&map_req_thread, NULL,
> &iwpm_mapping_reqs_handler, NULL);
> +	ret = pthread_create(&map_req_thread, NULL,
> iwpm_mapping_reqs_handler, NULL);
>  	if (ret)
>  		goto error_exit;
> 
> -	ret = pthread_create(&pending_msg_thread, NULL,
> &iwpm_pending_msgs_handler, NULL);
> +	ret = pthread_create(&pending_msg_thread, NULL,
> iwpm_pending_msgs_handler, NULL);
>  	if (ret)
>  		goto error_exit;

How did this work before?

Reviewed-by: Steve Wise <swise-7bPotxP6k4+P2YhJcF5u+vpXobYPEAuW@public.gmane.org>

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH 12/32] iwpmd: fix the prototype for the pthread_create start functions

[Jason Gunthorpe]

On Thu, Oct 13, 2016 at 02:09:44PM -0500, Steve Wise wrote:
> > @@ -1438,11 +1438,11 @@ int main(int argc, char *argv[])
> >  	pthread_cond_init(&cond_req_complete, NULL);
> >  	pthread_cond_init(&cond_pending_msg, NULL);
> > 
> > -	ret = pthread_create(&map_req_thread, NULL,
> > &iwpm_mapping_reqs_handler, NULL);
> > +	ret = pthread_create(&map_req_thread, NULL,
> > iwpm_mapping_reqs_handler, NULL);
> >  	if (ret)
> >  		goto error_exit;
> > 
> > -	ret = pthread_create(&pending_msg_thread, NULL,
> > &iwpm_pending_msgs_handler, NULL);
> > +	ret = pthread_create(&pending_msg_thread, NULL,
> > iwpm_pending_msgs_handler, NULL);
> >  	if (ret)
> >  		goto error_exit;
> 
> How did this work before?

gcc will not warn if an old-style prototype eg () is passed in to a
typed function pointer.

An earlier patch in the series changed it to (void) which broke the
compile.. I reorderd things to avoid this in the PR.

Jason
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 13/32] librdmacm: mark symbols static where possible

[Christoph Hellwig]

Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
---
 librdmacm/addrinfo.c              |  2 +-
 librdmacm/examples/rcopy.c        |  4 ++--
 librdmacm/examples/rdma_client.c  | 10 +++++-----
 librdmacm/examples/rdma_server.c  | 10 +++++-----
 librdmacm/examples/rdma_xclient.c | 10 +++++-----
 librdmacm/examples/rdma_xserver.c | 10 +++++-----
 librdmacm/rsocket.c               |  2 +-
 7 files changed, 24 insertions(+), 24 deletions(-)

diff --git a/librdmacm/addrinfo.c b/librdmacm/addrinfo.c
index 2f4e674..2c515a2 100644
--- a/librdmacm/addrinfo.c
+++ b/librdmacm/addrinfo.c
@@ -43,7 +43,7 @@
 #include <rdma/rdma_cma.h>
 #include <infiniband/ib.h>
 
-struct rdma_addrinfo nohints;
+static struct rdma_addrinfo nohints;
 
 static void ucma_convert_to_ai(struct addrinfo *ai, struct rdma_addrinfo *rai)
 {
diff --git a/librdmacm/examples/rcopy.c b/librdmacm/examples/rcopy.c
index 085b017..38f6929 100644
--- a/librdmacm/examples/rcopy.c
+++ b/librdmacm/examples/rcopy.c
@@ -59,8 +59,8 @@ static char *src_file;
 static struct timeval start, end;
 //static void buf[1024 * 1024];
 static uint64_t bytes;
-int fd;
-void *file_addr;
+static int fd;
+static void *file_addr;
 
 enum {
 	CMD_NOOP,
diff --git a/librdmacm/examples/rdma_client.c b/librdmacm/examples/rdma_client.c
index d7f65f5..fd7bbc2 100644
--- a/librdmacm/examples/rdma_client.c
+++ b/librdmacm/examples/rdma_client.c
@@ -39,11 +39,11 @@
 static char *server = "127.0.0.1";
 static char *port = "7471";
 
-struct rdma_cm_id *id;
-struct ibv_mr *mr, *send_mr;
-int send_flags;
-uint8_t send_msg[16];
-uint8_t recv_msg[16];
+static struct rdma_cm_id *id;
+static struct ibv_mr *mr, *send_mr;
+static int send_flags;
+static uint8_t send_msg[16];
+static uint8_t recv_msg[16];
 
 static int run(void)
 {
diff --git a/librdmacm/examples/rdma_server.c b/librdmacm/examples/rdma_server.c
index d98d11a..3c7c15e 100644
--- a/librdmacm/examples/rdma_server.c
+++ b/librdmacm/examples/rdma_server.c
@@ -38,11 +38,11 @@
 
 static char *port = "7471";
 
-struct rdma_cm_id *listen_id, *id;
-struct ibv_mr *mr, *send_mr;
-int send_flags;
-uint8_t send_msg[16];
-uint8_t recv_msg[16];
+static struct rdma_cm_id *listen_id, *id;
+static struct ibv_mr *mr, *send_mr;
+static int send_flags;
+static uint8_t send_msg[16];
+static uint8_t recv_msg[16];
 
 static int run(void)
 {
diff --git a/librdmacm/examples/rdma_xclient.c b/librdmacm/examples/rdma_xclient.c
index 8dba266..d86fee1 100644
--- a/librdmacm/examples/rdma_xclient.c
+++ b/librdmacm/examples/rdma_xclient.c
@@ -40,12 +40,12 @@
 static char *server = "127.0.0.1";
 static char port[6] = "7471";
 
-struct rdma_cm_id *id;
-struct ibv_mr *mr;
-struct rdma_addrinfo hints;
+static struct rdma_cm_id *id;
+static struct ibv_mr *mr;
+static struct rdma_addrinfo hints;
 
-uint8_t send_msg[16];
-uint32_t srqn;
+static uint8_t send_msg[16];
+static uint32_t srqn;
 
 static int post_send(void)
 {
diff --git a/librdmacm/examples/rdma_xserver.c b/librdmacm/examples/rdma_xserver.c
index 69f170b..f838e7d 100644
--- a/librdmacm/examples/rdma_xserver.c
+++ b/librdmacm/examples/rdma_xserver.c
@@ -40,12 +40,12 @@
 
 static char *port = "7471";
 
-struct rdma_cm_id *listen_id, *id;
-struct ibv_mr *mr;
-struct rdma_addrinfo hints;
+static struct rdma_cm_id *listen_id, *id;
+static struct ibv_mr *mr;
+static struct rdma_addrinfo hints;
 
-uint8_t recv_msg[16];
-uint32_t srqn;
+static uint8_t recv_msg[16];
+static uint32_t srqn;
 
 static int create_srq(void)
 {
diff --git a/librdmacm/rsocket.c b/librdmacm/rsocket.c
index 7e7a38e..e0755dd 100644
--- a/librdmacm/rsocket.c
+++ b/librdmacm/rsocket.c
@@ -505,7 +505,7 @@ static int rs_scale_to_value(int value, int bits)
 		(void) rc;                                                     \
 	}
 
-void rs_configure(void)
+static void rs_configure(void)
 {
 	FILE *f;
 	static int init;
-- 
2.1.4

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 14/32] cxgb4: mark symbols static where possible

[Christoph Hellwig]

Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
---
 providers/cxgb4/dev.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/providers/cxgb4/dev.c b/providers/cxgb4/dev.c
index e1486ec..69b04ac 100644
--- a/providers/cxgb4/dev.c
+++ b/providers/cxgb4/dev.c
@@ -49,7 +49,7 @@
  * Macros needed to support the PCI Device ID Table ...
  */
 #define CH_PCI_DEVICE_ID_TABLE_DEFINE_BEGIN \
-	struct { \
+	static struct { \
 		unsigned vendor; \
 		unsigned device; \
 	} hca_table[] = {
-- 
2.1.4

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

RE: [PATCH 14/32] cxgb4: mark symbols static where possible

[Steve Wise]

> 
> Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>

Looks fine

Reviewed-by: Steve Wise <swise-7bPotxP6k4+P2YhJcF5u+vpXobYPEAuW@public.gmane.org>

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 15/32] cxgb4: remove unused function dump_wqe

[Christoph Hellwig]

Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
---
 providers/cxgb4/qp.c | 14 --------------
 1 file changed, 14 deletions(-)

diff --git a/providers/cxgb4/qp.c b/providers/cxgb4/qp.c
index 95b459a..376a00a 100644
--- a/providers/cxgb4/qp.c
+++ b/providers/cxgb4/qp.c
@@ -267,20 +267,6 @@ static int build_rdma_recv(struct c4iw_qp *qhp, union t4_recv_wr *wqe,
 	return 0;
 }
 
-void dump_wqe(void *arg)
-{
-	u64 *p = arg;
-	int len16;
-
-	len16 = be64_to_cpu(*p) & 0xff;
-	while (len16--) {
-		printf("%02x: %016llx ", (u8)(unsigned long)p, (long long)be64_to_cpu(*p));
-		p++;
-		printf("%016llx\n", (long long)be64_to_cpu(*p));
-		p++;
-	}
-}
-
 static void ring_kernel_db(struct c4iw_qp *qhp, u32 qid, u16 idx)
 {
 	struct ibv_modify_qp cmd;
-- 
2.1.4

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

RE: [PATCH 15/32] cxgb4: remove unused function dump_wqe

[Steve Wise]

> Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>

Looks fine.

Reviewed-by: Steve Wise <swise-7bPotxP6k4+P2YhJcF5u+vpXobYPEAuW@public.gmane.org>


--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: trivial sparse fixes for rdma-core

[Jason Gunthorpe]

On Thu, Oct 13, 2016 at 07:51:03PM +0200, Christoph Hellwig wrote:
> This is the low hanging fruit in terms of reducing sparse noise before
> looking into endianess and iomem annotations.

Looks fine to me. Since I gave them a good compile test I also
prepared the pull request for the first 15 that made it to the list:

https://github.com/linux-rdma/rdma-core/pull/19

Here is the patch to turn on gcc -Wstrict-prototypes

https://github.com/jgunthorpe/rdma-plumbing/commit/dd80c3b362ae90d0de36a6249409c20581bae864

I'll grab the rest of the patches into this PR when they make it to
the list.

Made one tiny change, re-orderd 'iwpmd: fix the prototype for the
pthread_create start functionss' to be before 'iwpmd: use proper
ANSI-C prototypes for functions without arguments' as this ordering
keeps every patch buildable.

Regards,
Jason
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: trivial sparse fixes for rdma-core

[Leon Romanovsky]

[-- Attachment #1: Type: text/plain, Size: 735 bytes --]

On Thu, Oct 13, 2016 at 01:03:26PM -0600, Jason Gunthorpe wrote:
> On Thu, Oct 13, 2016 at 07:51:03PM +0200, Christoph Hellwig wrote:
> > This is the low hanging fruit in terms of reducing sparse noise before
> > looking into endianess and iomem annotations.
>
> Looks fine to me. Since I gave them a good compile test I also
> prepared the pull request for the first 15 that made it to the list:
>
> https://github.com/linux-rdma/rdma-core/pull/19
>
> Here is the patch to turn on gcc -Wstrict-prototypes
>
> https://github.com/jgunthorpe/rdma-plumbing/commit/dd80c3b362ae90d0de36a6249409c20581bae864

I didn't look on it yet, but maybe someone already knows the answer.
Can we run static analyzer tools as part of our TravisCI runs?

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

Re: trivial sparse fixes for rdma-core

[Leon Romanovsky]

On Fri, Oct 14, 2016 at 7:37 AM, Leon Romanovsky <leon-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org> wrote:
> On Thu, Oct 13, 2016 at 01:03:26PM -0600, Jason Gunthorpe wrote:
>> On Thu, Oct 13, 2016 at 07:51:03PM +0200, Christoph Hellwig wrote:
>> > This is the low hanging fruit in terms of reducing sparse noise before
>> > looking into endianess and iomem annotations.
>>
>> Looks fine to me. Since I gave them a good compile test I also
>> prepared the pull request for the first 15 that made it to the list:
>>
>> https://github.com/linux-rdma/rdma-core/pull/19
>>
>> Here is the patch to turn on gcc -Wstrict-prototypes
>>
>> https://github.com/jgunthorpe/rdma-plumbing/commit/dd80c3b362ae90d0de36a6249409c20581bae864
>
> I didn't look on it yet, but maybe someone already knows the answer.
> Can we run static analyzer tools as part of our TravisCI runs?

It looks like the answer is yes:
https://developer.fedoraproject.org/tech/languages/c/static-analysis.html

Thanks
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: trivial sparse fixes for rdma-core

[Christoph Hellwig]

On Fri, Oct 14, 2016 at 07:37:05AM +0300, Leon Romanovsky wrote:
> I didn't look on it yet, but maybe someone already knows the answer.
> Can we run static analyzer tools as part of our TravisCI runs?

Just run CC=cgcc cmake for sparse..
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: trivial sparse fixes for rdma-core

[Jason Gunthorpe]

On Fri, Oct 14, 2016 at 07:18:43AM +0200, Christoph Hellwig wrote:
> On Fri, Oct 14, 2016 at 07:37:05AM +0300, Leon Romanovsky wrote:
> > I didn't look on it yet, but maybe someone already knows the answer.
> > Can we run static analyzer tools as part of our TravisCI runs?
> 
> Just run CC=cgcc cmake for sparse..

We need to run sparse in a pass/fail way for Travis, I use -Werror to
do this for gcc.

The sparse version easially installable in travis is 0.4.5~rc1, if we
need newer then we have to find an APT repository with it built for
trusty.

Jason
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: trivial sparse fixes for rdma-core

[Leon Romanovsky]

[-- Attachment #1: Type: text/plain, Size: 435 bytes --]

On Thu, Oct 13, 2016 at 07:51:03PM +0200, Christoph Hellwig wrote:
> This is the low hanging fruit in terms of reducing sparse noise before
> looking into endianess and iomem annotations.

Thanks, applied.

>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
> the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 819 bytes --]