On Tue, 2017-07-11 at 23:04 +0300, Dan Carpenter wrote:
> This is a semi-automatic email about new static checker warnings.
>
> The patch 75f141aaf48e: "target/tcm_loop: Use target_submit_tmr()
> instead of open-coding this function" from May 23, 2017, leads to the
> following Smatch complaint:
>
> drivers/target/loopback/tcm_loop.c:247 tcm_loop_issue_tmr()
> warn: variable dereferenced before check 'se_cmd' (see line 234)
>
> drivers/target/loopback/tcm_loop.c
> 223 tl_cmd = kmem_cache_zalloc(tcm_loop_cmd_cache, GFP_KERNEL);
> 224 if (!tl_cmd) {
> 225 pr_err("Unable to allocate memory for tl_cmd\n");
> 226 return ret;
> 227 }
> 228
> 229 init_completion(&tl_cmd->tmr_done);
> 230
> 231 se_cmd = &tl_cmd->tl_se_cmd;
> 232 se_sess = tl_tpg->tl_nexus->se_sess;
> 233
> 234 rc = target_submit_tmr(se_cmd, se_sess, tl_cmd->tl_sense_buf, lun,
> ^^^^^^
> Dereferenced here.
>
> 235 NULL, tmr, GFP_KERNEL, task,
> 236 TARGET_SCF_ACK_KREF);
> 237 if (rc < 0)
> 238 goto release;
> 239 wait_for_completion(&tl_cmd->tmr_done);
> 240 ret = se_cmd->se_tmr_req->response;
> 241 target_put_sess_cmd(se_cmd);
> 242
> 243 out:
> 244 return ret;
> 245
> 246 release:
> 247 if (se_cmd)
> 248 transport_generic_free_cmd(se_cmd, 0);
> 249 else
> 250 kmem_cache_free(tcm_loop_cmd_cache, tl_cmd);
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> This is dead code.
>
> 251 goto out;
> 252 }
Hello Dan,
Although I agree that the code marked as dead code is indeed dead code, I
think that this code became dead code due to another patch, namely commit
4d3895d5ea43 ("target/tcm_loop: Merge struct tcm_loop_cmd and struct
tcm_loop_tmr"). Anyway, thanks for the report. Removing the dead code
would be good but since this warning does not indicate a bug removing the
dead code is not urgent.
Bart.--
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
