If we cannot setup a cmd because we run out of ring space or global pages release the blocks before sleeping. This prevents a deadlock where dev0 has waiting_blocks set and needs N blocks, but dev1 to devX have each allocated N / X blocks and also hit the global block limit so they went to sleep.
find_free_blocks is not able to take the sleeping dev's blocks becaause their waiting_blocks is set and even if it was not the block returned by find_last_bit could equal dbi_max. The latter will probably never happen because DATA_BLOCK_BITS is so high but in the next patches DATA_BLOCK_BITS and TCMU_GLOBAL_MAX_BLOCKS will be settable so it might be lower and could happen. Signed-off-by: Mike Christie <[email protected]> --- drivers/target/target_core_user.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/drivers/target/target_core_user.c b/drivers/target/target_core_user.c index 5226b82..612ac65 100644 --- a/drivers/target/target_core_user.c +++ b/drivers/target/target_core_user.c @@ -806,6 +806,13 @@ static int tcmu_setup_cmd_timer(struct tcmu_cmd *tcmu_cmd) int ret; DEFINE_WAIT(__wait); + /* + * Don't leave commands partially setup because the unmap + * thread might need the blocks to make forward progress. + */ + tcmu_cmd_free_data(tcmu_cmd, tcmu_cmd->dbi_cur); + tcmu_cmd_reset_dbi_cur(tcmu_cmd); + prepare_to_wait(&udev->wait_cmdr, &__wait, TASK_INTERRUPTIBLE); pr_debug("sleeping for ring space\n"); -- 1.8.3.1
