4.9-stable review patch. If anyone has any objections, please let me know.
------------------
From: Sameer Wadgaonkar sameer.wadgaonkar@unisys.com
[ Upstream commit 3c2bf0bd08123f3497bd3e84bd9088c937b0cb40 ]
The root issue is that we are not allowed to have items on the stack being passed to "DMA" like operations. In this case we have a vmcall and an inline completion of scsi command.
This patch fixes the issue by moving the variables on stack in do_scsi_nolinuxstat() to heap memory.
Signed-off-by: Sameer Wadgaonkar sameer.wadgaonkar@unisys.com Signed-off-by: David Kershner david.kershner@unisys.com Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org Signed-off-by: Sasha Levin alexander.levin@microsoft.com Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org --- drivers/staging/unisys/visorhba/visorhba_main.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-)
--- a/drivers/staging/unisys/visorhba/visorhba_main.c +++ b/drivers/staging/unisys/visorhba/visorhba_main.c @@ -842,7 +842,7 @@ static void do_scsi_nolinuxstat(struct uiscmdrsp *cmdrsp, struct scsi_cmnd *scsicmd) { struct scsi_device *scsidev; - unsigned char buf[36]; + unsigned char *buf; struct scatterlist *sg; unsigned int i; char *this_page; @@ -857,6 +857,10 @@ do_scsi_nolinuxstat(struct uiscmdrsp *cm if (cmdrsp->scsi.no_disk_result == 0) return;
+ buf = kzalloc(sizeof(char) * 36, GFP_KERNEL); + if (!buf) + return; + /* Linux scsi code wants a device at Lun 0 * to issue report luns, but we don't want * a disk there so we'll present a processor @@ -868,6 +872,7 @@ do_scsi_nolinuxstat(struct uiscmdrsp *cm if (scsi_sg_count(scsicmd) == 0) { memcpy(scsi_sglist(scsicmd), buf, cmdrsp->scsi.bufflen); + kfree(buf); return; }
@@ -879,6 +884,7 @@ do_scsi_nolinuxstat(struct uiscmdrsp *cm memcpy(this_page, buf + bufind, sg[i].length); kunmap_atomic(this_page_orig); } + kfree(buf); } else { devdata = (struct visorhba_devdata *)scsidev->host->hostdata; for_each_vdisk_match(vdisk, devdata, scsidev) {