- 01 Mar, 2011 7 commits
-
-
Dan Carpenter authored
Negate has higher precedence than bitwise AND. FCPHF_CRC_UNCHECKED is 0x1 so the original code is equivalent to: if (!fr_flags(fp)) { ... Signed-off-by:Dan Carpenter <error27@gmail.com> Signed-off-by:
Robert Love <robert.w.love@intel.com> Signed-off-by:
James Bottomley <James.Bottomley@suse.de>
-
Bhanu Prakash Gollapudi authored
When L2 driver is unloaded, libfcoe_destroy tries to access the fcoe transport structure matching the netdev. However, since the netdev is unregistered by that time, it fails to do so. Hence the stale mappings exists in the fcoe-netdev list. Handle NETDEV_UREGISTER device notification mechanism to remove the stale fcoe-netdev mapping. Signed-off-by:
Bhanu Prakash Gollapudi <bprakash@broadcom.com> Signed-off-by:
-
Bhanu Prakash Gollapudi authored
both fcoe and bnx2fc drivers can access the common definition of FCOE_MTU. Signed-off-by:
-
Bhanu Prakash Gollapudi authored
fc_fill_fc_hdr() expects fc_frame as an argument. Introduce __fc_fill_fc_hdr to accept fc_frame_header as an argument. Signed-off-by:
-
Vasu Dev authored
EM anchors list initialization for only master port was not enough to keep npiv working as described here:- https://lists.open-fcoe.org/pipermail/devel/2011-January/011063.html So this patch moves fc_exch_mgr_list_clone to update npiv ports EMs once EM anchors list initialized. Also some cleanup, no need to set lport = NULL as that always get initialized later. Signed-off-by:
Vasu Dev <vasu.dev@intel.com> Signed-off-by:
-
Parikh, Neerav authored
When abort for an exchange timed out it didn't release the reference to the exchange resulting in a memory leak. After discussion with the author of the patch (CC) that introduced this bug it was suggested to revert that patch. This reverts commit ea3e2e72. Signed-off by: Neerav Parikh <Neerav.Parikh@intel.com> Tested-by:
Ross Brattain <ross.b.brattain@intel.com> Signed-off-by:
-
Parikh, Neerav authored
When an fcoe interface is being destroyed; in the process the fcoe driver will try to release all the resources it had allocated for that interface including rports. But, it seems that it does not release the reference held for the name server rport in that process resulting into a memory leak. This patch fixes that memory leak. Signed-off-by:
Neerav Parikh <neerav.parikh@intel.com> Tested-by:
-
- 28 Feb, 2011 16 commits
-
-
adam radford authored
Signed-off-by:
Adam Radford <aradford@gmail.com> Signed-off-by:
-
adam radford authored
Signed-off-by:
-
adam radford authored
Signed-off-by:
-
adam radford authored
The following patch for megaraid_sas fixes fault state handling in megasas_transition_to_ready(). Signed-off-by:
-
adam radford authored
The following patch fixes an incorrect tasklet_init() call in megasas_init_fw() to use instancet->tasklet. Signed-off-by:
-
adam radford authored
The following patch for megaraid_sas adds a missing check for MR_EVT_CFG_CLEARED in megasas_aen_polling(). Signed-off-by:
-
adam radford authored
Signed-off-by:
-
adam radford authored
Signed-off-by:
-
adam radford authored
The following patch for megaraid_sas fixes megasas_probe_one() to clear MSI-X flags in kdump when the 'reset_devices' kernel parameter is passed in. Signed-off-by:
-
adam radford authored
The following patch for megaraid_sas calls tasklet_schedule() even if outbound_intr_status == 0 for MFI based boards in MSI-X mode. Signed-off-by:
-
adam radford authored
Signed-off-by:
-
adam radford authored
Signed-off-by:
-
adam radford authored
Signed-off-by:
-
jack_wang authored
In some test envirenment, there is loopback topology test. We should handle this during discovery. Signed-off-by:
Jack Wang <jack_wang@usish.com> Signed-off-by:
-
Mike Christie authored
Instead of blocking the entire host when the port's queueing limit is hit, we should only block the port's target. This will allow IO to other ports to execute. Signed-off-by:
Mike Christie <michaelc@cs.wisc.edu> Acked-by:
James Smart <james.smart@emulex.com> Signed-off-by:
-
Mike Christie authored
If the port takes a while to transition we could exhaust the retries when using DID_TRANSPORT_DISRUPTED. For this case we do not want to use any of the cmd's retries, because if the command was running then when it got failed the retry counter was already incremented. And if this is the first time we are seeing the command, (it got queued because it slipped through during the race) then it should not have its retries incremented. The fc class will decide the correct handling later. Signed-off-by:
-
- 25 Feb, 2011 17 commits
-
-
Madhuranath Iyengar authored
We'll be using the new version number scheme from now on. The 'k' in the version in the past will be removed. The format will be: <major>.<minor>.<sub_minor>.<iter> A scsi-misc submission increments <sub_minor> and resets <iter>. An scsi-rc-fixes submission, increments <iter>. Signed-off-by:
Madhuranath Iyengar <Madhu.Iyengar@qlogic.com> Signed-off-by:
-
Andrew Vasquez authored
The fc_block_scsi_eh() might return with status FAST_IO_FAIL indicating I/O has been terminated due to fast_io_fail timeout. In this case the rport is still blocked, so any error recovery will be failing on this port. Hence we need to check if the return value from fc_block_scsi_eh() is something other than 0, in which case it should just return with that status. Signed-off-by:
Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by:
-
Saurav Kashyap authored
If adapter is not online, the driver will not process the response queue, even on getting an interrupt. Signed-off-by:
Saurav Kashyap <saurav.kashyap@qlogic.com> Signed-off-by:
-
Andrew Vasquez authored
Signed-off-by:
-
Andrew Vasquez authored
Based on reading of the FCP2/4 specification, the driver cannot expect the storage device to send FCP_RSP information. Instead, the driver should interpret the data which is present in the frame to base decisions on the success or failure of the system. Signed-off-by:
-
Andrew Vasquez authored
By not clearing the 'login needed' nor 'login outstanding' states for an fcport after a big-hammer (adapter reset), the driver may not properly perform a PLOGI/PRLI sequence (and lose visibility to the rport) during a follow-on SNS scan. Signed-off-by:
-
Andrew Vasquez authored
Signed-off-by:
Giridhar Malavali <giridhar.malavali@qlogic.com> Signed-off-by:
-
Giridhar Malavali authored
Signed-off-by:
-
Giridhar Malavali authored
Signed-off-by:
-
Giridhar Malavali authored
Signed-off-by:
-
Giridhar Malavali authored
Signed-off-by:
-
Chad Dupuis authored
The additional increment of dsd_seg was causing the compiler to throw an array out of bounds warning. This patch moves to a direct assignment of cmd_pkt->fcp_data_dseg_len so that the compiler doesn't generate an array out bounds warning. Signed-off-by:
Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by:
-
Madhuranath Iyengar authored
Also in qla_os.c, rename the function, remove DEF_SCSI_QCMD, etc. Signed-off-by:
-
Christof Schmitt authored
Query the FC symbolic port name for reporting in the fc_host sysfs and enable the symbolic_name attribute in the fc_host sysfs. When running in NPIV mode, extend the symbolic port name with the devno and the hostname. This allows better identification of Linux systems for SAN and storage administrators. Signed-off-by:
Christof Schmitt <christof.schmitt@de.ibm.com> Signed-off-by:
Steffen Maier <maier@linux.vnet.ibm.com> Signed-off-by:
-
Christof Schmitt authored
Add request code and corresponding structs for "Get Symbolic Port Name" (GSPN_ID) request. Signed-off-by:
-
Christof Schmitt authored
The SCSI host and transport templates are the only members left in the global zfcp_data struct. Move them out of zfcp_data and remove the now unused zfcp_data struct. Also update the names of the register and unregister functions to use the zfcp_scsi prefix. Signed-off-by:
-
Christof Schmitt authored
For task management commands, only LUN and flags are required. The regular FCP setup already sets the LUN in the fcp_cmnd. All is required for merging the function is setting up the TM flags. Signed-off-by:
-
