The
problem: When
a SCSI device attached to a device handler is deleted, userland processes
currently performing I/O on the device will I/O hang forever. The
root cause: When
scsi_dh_activate returns SCSI_DH_NOSYS the activate_complete callback is not
called and the error is not propagated to the multipath
layer. A similar situation occurs when the handler is in the process of being
deleted (e.g. the SCSI device is in a SDEV_CANCEL or SDEV_DEL state). Attached
is the SCSI H/W handler part of this patch. Menny Hamburger Engineer Dell | IDC office +972 97698789, fax +972 97698889 Dell IDC. 4 Hacharoshet St, Raanana 43657, Israel |
Attachment:
scsi-scsi_dh-handle_device-deletion.patch
Description: scsi-scsi_dh-handle_device-deletion.patch
-- dm-devel mailing list dm-devel@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/dm-devel