[PATCH] RDMA/srp (gcc13): force int types for max_send_sge and can_queue

[PATCH] RDMA/srp (gcc13): force int types for max_send_sge and can_queue

[Jiri Slaby (SUSE)]

Since gcc13, each member of an enum has the same type as the enum [1]. And
that is inherited from its members. Provided "SRP_TAG_TSK_MGMT = 1U << 31",
SRP_MAX_SGE and SRP_TSK_MGMT_SQ_SIZE are unsigned ints.

This results in the following warnings:
  include/linux/minmax.h:20:35: error: comparison of distinct pointer types lacks a cast
  drivers/infiniband/ulp/srp/ib_srp.c:563:42: note: in expansion of macro 'min'

  include/linux/minmax.h:20:35: error: comparison of distinct pointer types lacks a cast
  drivers/infiniband/ulp/srp/ib_srp.c:2369:27: note: in expansion of macro 'min'

Force the use of min_t() instead of min() to use int for all those, as
this is what both targets
  target->scsi_host->can_queue
and
  init_attr->cap.max_send_sge
are.

[1] https://gcc.gnu.org/bugzilla/show_bug.cgi?id=36113

Cc: Martin Liska <mliska@suse.cz>
Cc: Bart Van Assche <bvanassche@acm.org>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Leon Romanovsky <leon@kernel.org>
Cc: linux-rdma@vger.kernel.org
Signed-off-by: Jiri Slaby (SUSE) <jirislaby@kernel.org>
---
 drivers/infiniband/ulp/srp/ib_srp.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c
index 1075c2ac8fe2..7db487da8293 100644
--- a/drivers/infiniband/ulp/srp/ib_srp.c
+++ b/drivers/infiniband/ulp/srp/ib_srp.c
@@ -560,7 +560,8 @@ static int srp_create_ch_ib(struct srp_rdma_ch *ch)
 	init_attr->cap.max_send_wr     = m * target->queue_size;
 	init_attr->cap.max_recv_wr     = target->queue_size + 1;
 	init_attr->cap.max_recv_sge    = 1;
-	init_attr->cap.max_send_sge    = min(SRP_MAX_SGE, attr->max_send_sge);
+	init_attr->cap.max_send_sge    = min_t(int, SRP_MAX_SGE,
+			attr->max_send_sge);
 	init_attr->sq_sig_type         = IB_SIGNAL_REQ_WR;
 	init_attr->qp_type             = IB_QPT_RC;
 	init_attr->send_cq             = send_cq;
@@ -2366,7 +2367,7 @@ static void srp_cm_rep_handler(struct ib_cm_id *cm_id,
 		 * bounce requests back to the SCSI mid-layer.
 		 */
 		target->scsi_host->can_queue
-			= min(ch->req_lim - SRP_TSK_MGMT_SQ_SIZE,
+			= min_t(int, ch->req_lim - SRP_TSK_MGMT_SQ_SIZE,
 			      target->scsi_host->can_queue);
 		target->scsi_host->cmd_per_lun
 			= min_t(int, target->scsi_host->can_queue,
-- 
2.38.1

Re: [PATCH] RDMA/srp (gcc13): force int types for max_send_sge and can_queue

[Bart Van Assche]

On 10/31/22 04:45, Jiri Slaby (SUSE) wrote:
> Since gcc13, each member of an enum has the same type as the enum [1]. And
> that is inherited from its members. Provided "SRP_TAG_TSK_MGMT = 1U << 31",
> SRP_MAX_SGE and SRP_TSK_MGMT_SQ_SIZE are unsigned ints.
> 
> This results in the following warnings:
>    include/linux/minmax.h:20:35: error: comparison of distinct pointer types lacks a cast
>    drivers/infiniband/ulp/srp/ib_srp.c:563:42: note: in expansion of macro 'min'
> 
>    include/linux/minmax.h:20:35: error: comparison of distinct pointer types lacks a cast
>    drivers/infiniband/ulp/srp/ib_srp.c:2369:27: note: in expansion of macro 'min'
> 
> Force the use of min_t() instead of min() to use int for all those, as
> this is what both targets
>    target->scsi_host->can_queue
> and
>    init_attr->cap.max_send_sge
> are.
> 
> [1] https://gcc.gnu.org/bugzilla/show_bug.cgi?id=36113
> 
> Cc: Martin Liska <mliska@suse.cz>
> Cc: Bart Van Assche <bvanassche@acm.org>
> Cc: Jason Gunthorpe <jgg@ziepe.ca>
> Cc: Leon Romanovsky <leon@kernel.org>
> Cc: linux-rdma@vger.kernel.org
> Signed-off-by: Jiri Slaby (SUSE) <jirislaby@kernel.org>
> ---
>   drivers/infiniband/ulp/srp/ib_srp.c | 5 +++--
>   1 file changed, 3 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c
> index 1075c2ac8fe2..7db487da8293 100644
> --- a/drivers/infiniband/ulp/srp/ib_srp.c
> +++ b/drivers/infiniband/ulp/srp/ib_srp.c
> @@ -560,7 +560,8 @@ static int srp_create_ch_ib(struct srp_rdma_ch *ch)
>   	init_attr->cap.max_send_wr     = m * target->queue_size;
>   	init_attr->cap.max_recv_wr     = target->queue_size + 1;
>   	init_attr->cap.max_recv_sge    = 1;
> -	init_attr->cap.max_send_sge    = min(SRP_MAX_SGE, attr->max_send_sge);
> +	init_attr->cap.max_send_sge    = min_t(int, SRP_MAX_SGE,
> +			attr->max_send_sge);
>   	init_attr->sq_sig_type         = IB_SIGNAL_REQ_WR;
>   	init_attr->qp_type             = IB_QPT_RC;
>   	init_attr->send_cq             = send_cq;
> @@ -2366,7 +2367,7 @@ static void srp_cm_rep_handler(struct ib_cm_id *cm_id,
>   		 * bounce requests back to the SCSI mid-layer.
>   		 */
>   		target->scsi_host->can_queue
> -			= min(ch->req_lim - SRP_TSK_MGMT_SQ_SIZE,
> +			= min_t(int, ch->req_lim - SRP_TSK_MGMT_SQ_SIZE,
>   			      target->scsi_host->can_queue);
>   		target->scsi_host->cmd_per_lun
>   			= min_t(int, target->scsi_host->can_queue,

I don't like this patch because it changes min() into min_t(). The 
former checks its arguments types while the latter uses casts and hence 
ignores most information that is present in the argument types. Please 
split the enum in ib_srp.h instead of changing min() into min_t().

Thanks,

Bart.