On 9/20/2017 12:18 AM, Jason Gunthorpe wrote:
The match step determines if the provider should bind to the sysfs device,
and the allocation step creates the verbs_device and cleanly matches the
uninit_device step which frees it.
This split makes it simpler to factor out all the duplicated code in the
match step.
Signed-off-by: Jason Gunthorpe <jgunthorpe@xxxxxxxxxxxxxxxxxxxx>
---
libibverbs/driver.h | 21 ++++++++++++++-
libibverbs/init.c | 74 ++++++++++++++++++++++++++++++++++-------------------
2 files changed, 68 insertions(+), 27 deletions(-)
diff --git a/libibverbs/driver.h b/libibverbs/driver.h
index f9f5cfa2308c2f..397441f49a0e5a 100644
--- a/libibverbs/driver.h
+++ b/libibverbs/driver.h
@@ -40,6 +40,7 @@
#include <infiniband/kern-abi.h>
#include <ccan/list.h>
#include <config.h>
+#include <stdbool.h>
#ifdef __cplusplus
# define BEGIN_C_DECLS extern "C" {
@@ -95,10 +96,27 @@ struct verbs_qp {
struct verbs_xrcd *xrcd;
};
+/* A rdma device detected in sysfs */
+struct verbs_sysfs_dev {
+ struct list_node entry;
+ void *provider_data;
It looks as this field isn't used in the series, correct ? if this is
the case better add when it'll become applicable.
+ char sysfs_name[IBV_SYSFS_NAME_MAX];
+ char ibdev_name[IBV_SYSFS_NAME_MAX];
+ char sysfs_path[IBV_SYSFS_PATH_MAX];
+ char ibdev_path[IBV_SYSFS_PATH_MAX];
+ int abi_ver;
+ struct timespec time_created;
+};
+
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
