From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Nicholas A. Bellinger" Subject: Re: linux-next: build failure after merge of the target-updates tree Date: Mon, 17 Sep 2012 17:57:24 -0700 Message-ID: <1347929844.10711.136.camel@haakon2.linux-iscsi.org> References: <20120917134142.c7948ae04bdeabe1a863e464@canb.auug.org.au> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20120917134142.c7948ae04bdeabe1a863e464@canb.auug.org.au> Sender: target-devel-owner@vger.kernel.org To: Stephen Rothwell Cc: linux-next@vger.kernel.org, linux-kernel@vger.kernel.org, Roland Dreier , Paolo Bonzini , target-devel List-Id: linux-next.vger.kernel.org On Mon, 2012-09-17 at 13:41 +1000, Stephen Rothwell wrote: > Hi Nicholas, > > After merging the target-updates tree, today's linux-next build (x86_64 > allmodconfig) failed like this: > > drivers/target/target_core_transport.c: In function 'transport_get_sense_buffer': > drivers/target/target_core_transport.c:582:22: error: 'struct target_core_fabric_ops' has no member named 'set_fabric_sense_len' > > I can only assume that this has been caused by the (largish) changes that > went into Linus' tree recently conflicting with changes in the > target-updates tree. > > I have dropped the target-updates tree for today. Please sort the mess out. Hi Stephen, Whoops, sorry about that.. So target-pending/for-next has been rebased + now includes the following change to Roland's original patch that drops TFO->set_fabric_sense_len core+fabric usage for v3.7-rc1. As reported, this ended up conflicting with Paolo's recent pSCSI backend sense data handling bug-fixes that went in for v3.6-rc6. So with the following, for-next should be looking as expected now. Thanks! --nab diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c index fd0d0f0..d6d4844 100644 --- a/drivers/target/target_core_transport.c +++ b/drivers/target/target_core_transport.c @@ -567,9 +567,7 @@ static void target_complete_failure_work(struct work_struct *work) */ static unsigned char *transport_get_sense_buffer(struct se_cmd *cmd) { - unsigned char *buffer = cmd->sense_buffer; struct se_device *dev = cmd->se_dev; - u32 offset = 0; WARN_ON(!cmd->se_lun); @@ -579,14 +577,11 @@ static unsigned char *transport_get_sense_buffer(struct se_cmd *cmd) if (cmd->se_cmd_flags & SCF_SENT_CHECK_CONDITION) return NULL; - offset = cmd->se_tfo->set_fabric_sense_len(cmd, TRANSPORT_SENSE_BUFFER); - - /* Automatically padded */ - cmd->scsi_sense_length = TRANSPORT_SENSE_BUFFER + offset; + cmd->scsi_sense_length = TRANSPORT_SENSE_BUFFER; pr_debug("HBA_[%u]_PLUG[%s]: Requesting sense for SAM STATUS: 0x%02x\n", dev->se_hba->hba_id, dev->transport->name, cmd->scsi_status); - return &buffer[offset]; + return cmd->sense_buffer; } void target_complete_cmd(struct se_cmd *cmd, u8 scsi_status)