On Mon, 2011-11-28 at 04:07 -0800, Nicholas A. Bellinger wrote: > On Mon, 2011-11-28 at 12:50 +0100, Sebastian Andrzej Siewior wrote: > > On 11/28/2011 12:47 PM, Nicholas A. Bellinger wrote: > > > > > Note that IBLOCK and PSCSI are already using for_each_sg when walking > > > task->task_sg memory, and making FILEIO do this as well is what we > > > should be doing. However, I'm curious what your use-case was to trigger > > > this with fileio, as we don't expect a single task to ever use chained > > > sg in order to trigger this bug..? > > > > > > > This script: > > > > #!/bin/sh > > modprobe target_core_mod > > mount a -t configfs /sys/kernel/config > > > > CONFIGFS=/sys/kernel/config/ > > TARGET=/sys/kernel/config/target/core/ > > export FABRIC=/sys/kernel/config/target/loopback/ > > > > mkdir -p $TARGET/fileio_0/fileio > > echo "fd_dev_name=/root/file.bin,fd_dev_size=31457280" > > > $TARGET/fileio_0/fileio/control > > echo 1 > $TARGET/fileio_0/fileio/enable > > > > mkdir -p $FABRIC/naa.6001405c3214b06a/tpgt_1 > > mkdir $FABRIC/naa.6001405c3214b06a/tpgt_1/lun/lun_0 > > echo naa.6001405c3214b06b > $FABRIC/naa.6001405c3214b06a/tpgt_1/nexus > > > > ln -sv $TARGET/fileio_0/fileio > > $FABRIC/naa.6001405c3214b06a/tpgt_1/lun/lun_0/virtual_scsi_port > > > > > > followed by > > - cfdiks /dev/sda (make 1 partition) > > - mkfs.ext2 /dev/sda > > > > > Is this with a manually set small max_sectors fileio backend, or a > > > limited block_device max_sectors with fileio that would cause> 1 task > > > allocation to occur on your setup..? > > > > Just the script I mentioned plus the two commands afterwords. The crash > > happens during mkfs.ext2 > > > > Ok, so no special max_sectors with fileio. I'll take a look at this > with loopback LUNs later today. > Ok, I'm applying this patch to lio-core now as it addresses a real issue with fileio backends and loopback LUNs w/ per-chained SGL usage with SCSI LLD incoming struct scsi_cmnd scatterlist memory into target-core. This will need to be Cc'ed to stable as well for v3.1, and I'm adding a more detailed commit log for upstream. Thanks again for catching this! --nab -- 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
