On Fri, 2012-03-30 at 11:29 -0700, Roland Dreier wrote:
> From: Roland Dreier <roland@xxxxxxxxxxxxxxx>
>
> Now that all fabrics are converted over to using se_cmd->t_data_sg
> directly, we can drop the task sg chaining support. With the modern
> memory allocation in target core, task sg chaining is needless
> overhead -- we would split up the main cmd sglist into pieces, and
> then splice those pieces back together instead of just using the
> original list directly.
>
> Signed-off-by: Roland Dreier <roland@xxxxxxxxxxxxxxx>
> ---
Applied this patch minus the transport_allocate_data_tasks() specific
parts that conflict with hch's patch to drop task splitting.
Thanks Roland!
--nab
> drivers/target/target_core_transport.c | 89 +------------------------------
> include/target/target_core_base.h | 2 -
> include/target/target_core_fabric.h | 7 ---
> 3 files changed, 3 insertions(+), 95 deletions(-)
>
> diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c
> index e252925..58315b0 100644
> --- a/drivers/target/target_core_transport.c
> +++ b/drivers/target/target_core_transport.c
> @@ -3706,76 +3706,6 @@ static inline sector_t transport_limit_task_sectors(
> return sectors;
> }
>
> -
> -/*
> - * This function can be used by HW target mode drivers to create a linked
> - * scatterlist from all contiguously allocated struct se_task->task_sg[].
> - * This is intended to be called during the completion path by TCM Core
> - * when struct target_core_fabric_ops->check_task_sg_chaining is enabled.
> - */
> -void transport_do_task_sg_chain(struct se_cmd *cmd)
> -{
> - struct scatterlist *sg_first = NULL;
> - struct scatterlist *sg_prev = NULL;
> - int sg_prev_nents = 0;
> - struct scatterlist *sg;
> - struct se_task *task;
> - u32 chained_nents = 0;
> - int i;
> -
> - BUG_ON(!cmd->se_tfo->task_sg_chaining);
> -
> - /*
> - * Walk the struct se_task list and setup scatterlist chains
> - * for each contiguously allocated struct se_task->task_sg[].
> - */
> - list_for_each_entry(task, &cmd->t_task_list, t_list) {
> - if (!task->task_sg)
> - continue;
> -
> - if (!sg_first) {
> - sg_first = task->task_sg;
> - chained_nents = task->task_sg_nents;
> - } else {
> - sg_chain(sg_prev, sg_prev_nents, task->task_sg);
> - chained_nents += task->task_sg_nents;
> - }
> - /*
> - * For the padded tasks, use the extra SGL vector allocated
> - * in transport_allocate_data_tasks() for the sg_prev_nents
> - * offset into sg_chain() above.
> - *
> - * We do not need the padding for the last task (or a single
> - * task), but in that case we will never use the sg_prev_nents
> - * value below which would be incorrect.
> - */
> - sg_prev_nents = (task->task_sg_nents + 1);
> - sg_prev = task->task_sg;
> - }
> - /*
> - * Setup the starting pointer and total t_tasks_sg_linked_no including
> - * padding SGs for linking and to mark the end.
> - */
> - cmd->t_tasks_sg_chained = sg_first;
> - cmd->t_tasks_sg_chained_no = chained_nents;
> -
> - pr_debug("Setup cmd: %p cmd->t_tasks_sg_chained: %p and"
> - " t_tasks_sg_chained_no: %u\n", cmd, cmd->t_tasks_sg_chained,
> - cmd->t_tasks_sg_chained_no);
> -
> - for_each_sg(cmd->t_tasks_sg_chained, sg,
> - cmd->t_tasks_sg_chained_no, i) {
> -
> - pr_debug("SG[%d]: %p page: %p length: %d offset: %d\n",
> - i, sg, sg_page(sg), sg->length, sg->offset);
> - if (sg_is_chain(sg))
> - pr_debug("SG: %p sg_is_chain=1\n", sg);
> - if (sg_is_last(sg))
> - pr_debug("SG: %p sg_is_last=1\n", sg);
> - }
> -}
> -EXPORT_SYMBOL(transport_do_task_sg_chain);
> -
> /*
> * Break up cmd into chunks transport can handle
> */
> @@ -3830,7 +3760,7 @@ transport_allocate_data_tasks(struct se_cmd *cmd,
>
> for (i = 0; i < task_count; i++) {
> struct se_task *task;
> - unsigned int task_size, task_sg_nents_padded;
> + unsigned int task_size;
> struct scatterlist *sg;
> unsigned long flags;
> int count;
> @@ -3847,27 +3777,14 @@ transport_allocate_data_tasks(struct se_cmd *cmd,
> * in order to calculate the number per task SGL entries
> */
> task->task_sg_nents = DIV_ROUND_UP(task->task_size, PAGE_SIZE);
> - /*
> - * Check if the fabric module driver is requesting that all
> - * struct se_task->task_sg[] be chained together.. If so,
> - * then allocate an extra padding SG entry for linking and
> - * marking the end of the chained SGL for every task except
> - * the last one for (task_count > 1) operation, or skipping
> - * the extra padding for the (task_count == 1) case.
> - */
> - if (cmd->se_tfo->task_sg_chaining && (i < (task_count - 1))) {
> - task_sg_nents_padded = (task->task_sg_nents + 1);
> - } else
> - task_sg_nents_padded = task->task_sg_nents;
> -
> task->task_sg = kmalloc(sizeof(struct scatterlist) *
> - task_sg_nents_padded, GFP_KERNEL);
> + task->task_sg_nents, GFP_KERNEL);
> if (!task->task_sg) {
> cmd->se_dev->transport->free_task(task);
> return -ENOMEM;
> }
>
> - sg_init_table(task->task_sg, task_sg_nents_padded);
> + sg_init_table(task->task_sg, task->task_sg_nents);
>
> task_size = task->task_size;
>
> diff --git a/include/target/target_core_base.h b/include/target/target_core_base.h
> index 2c5c37e..0a09b85 100644
> --- a/include/target/target_core_base.h
> +++ b/include/target/target_core_base.h
> @@ -571,7 +571,6 @@ struct se_cmd {
> unsigned char *t_task_cdb;
> unsigned char __t_task_cdb[TCM_MAX_COMMAND_SIZE];
> unsigned long long t_task_lba;
> - u32 t_tasks_sg_chained_no;
> atomic_t t_fe_count;
> atomic_t t_se_count;
> atomic_t t_task_cdbs_left;
> @@ -592,7 +591,6 @@ struct se_cmd {
> struct completion t_transport_stop_comp;
> struct completion transport_lun_fe_stop_comp;
> struct completion transport_lun_stop_comp;
> - struct scatterlist *t_tasks_sg_chained;
>
> struct work_struct work;
>
> diff --git a/include/target/target_core_fabric.h b/include/target/target_core_fabric.h
> index 13dc827..cc64a94 100644
> --- a/include/target/target_core_fabric.h
> +++ b/include/target/target_core_fabric.h
> @@ -3,12 +3,6 @@
>
> struct target_core_fabric_ops {
> struct configfs_subsystem *tf_subsys;
> - /*
> - * Optional to signal struct se_task->task_sg[] padding entries
> - * for scatterlist chaining using transport_do_task_sg_link(),
> - * disabled by default
> - */
> - bool task_sg_chaining;
> char *(*get_fabric_name)(void);
> u8 (*get_fabric_proto_ident)(struct se_portal_group *);
> char *(*tpg_get_wwn)(struct se_portal_group *);
> @@ -124,7 +118,6 @@ int transport_generic_handle_cdb_map(struct se_cmd *);
> int transport_generic_handle_data(struct se_cmd *);
> int transport_generic_map_mem_to_cmd(struct se_cmd *cmd,
> struct scatterlist *, u32, struct scatterlist *, u32);
> -void transport_do_task_sg_chain(struct se_cmd *);
> int transport_generic_new_cmd(struct se_cmd *);
>
> void transport_generic_process_write(struct se_cmd *);
--
To unsubscribe from this list: send the line "unsubscribe target-devel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
