This is a note to let you know that I've just added the patch titled
xhci: Fix incorrect tracking of free space on transfer rings
to the 6.3-stable tree which can be found at:
http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary
The filename of the patch is:
xhci-fix-incorrect-tracking-of-free-space-on-transfer-rings.patch
and it can be found in the queue-6.3 subdirectory.
If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@xxxxxxxxxxxxxxx> know about it.
>From fe82f16aafdaf8002281d3b9524291d4a4a28460 Mon Sep 17 00:00:00 2001
From: Mathias Nyman <mathias.nyman@xxxxxxxxxxxxxxx>
Date: Mon, 15 May 2023 16:40:59 +0300
Subject: xhci: Fix incorrect tracking of free space on transfer rings
From: Mathias Nyman <mathias.nyman@xxxxxxxxxxxxxxx>
commit fe82f16aafdaf8002281d3b9524291d4a4a28460 upstream.
This incorrect tracking caused unnecessary ring expansion in some
usecases which over days of use consume a lot of memory.
xhci driver tries to keep track of free transfer blocks (TRBs) on the
ring buffer, but failed to add back some cancelled transfers that were
turned into no-op operations instead of just moving past them.
This can happen if there are several queued pending transfers which
then are cancelled in reverse order.
Solve this by counting the numer of steps we move the dequeue pointer
once we complete a transfer, and add it to the number of free trbs
instead of just adding the trb number of the current transfer.
This way we ensure we count the no-op trbs on the way as well.
Fixes: 55f6153d8cc8 ("xhci: remove extra loop in interrupt context")
Cc: stable@xxxxxxxxxxxxxxx
Reported-by: Miller Hunter <MillerH@xxxxxxxxxxxxxxx>
Closes: https://bugzilla.kernel.org/show_bug.cgi?id=217242
Tested-by: Miller Hunter <MillerH@xxxxxxxxxxxxxxx>
Signed-off-by: Mathias Nyman <mathias.nyman@xxxxxxxxxxxxxxx>
Link: https://lore.kernel.org/r/20230515134059.161110-3-mathias.nyman@xxxxxxxxxxxxxxx
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
---
drivers/usb/host/xhci-ring.c | 29 ++++++++++++++++++++++++++++-
1 file changed, 28 insertions(+), 1 deletion(-)
--- a/drivers/usb/host/xhci-ring.c
+++ b/drivers/usb/host/xhci-ring.c
@@ -276,6 +276,26 @@ static void inc_enq(struct xhci_hcd *xhc
trace_xhci_inc_enq(ring);
}
+static int xhci_num_trbs_to(struct xhci_segment *start_seg, union xhci_trb *start,
+ struct xhci_segment *end_seg, union xhci_trb *end,
+ unsigned int num_segs)
+{
+ union xhci_trb *last_on_seg;
+ int num = 0;
+ int i = 0;
+
+ do {
+ if (start_seg == end_seg && end >= start)
+ return num + (end - start);
+ last_on_seg = &start_seg->trbs[TRBS_PER_SEGMENT - 1];
+ num += last_on_seg - start;
+ start_seg = start_seg->next;
+ start = start_seg->trbs;
+ } while (i++ <= num_segs);
+
+ return -EINVAL;
+}
+
/*
* Check to see if there's room to enqueue num_trbs on the ring and make sure
* enqueue pointer will not advance into dequeue segment. See rules above.
@@ -2140,6 +2160,7 @@ static int finish_td(struct xhci_hcd *xh
u32 trb_comp_code)
{
struct xhci_ep_ctx *ep_ctx;
+ int trbs_freed;
ep_ctx = xhci_get_ep_ctx(xhci, ep->vdev->out_ctx, ep->ep_index);
@@ -2209,9 +2230,15 @@ static int finish_td(struct xhci_hcd *xh
}
/* Update ring dequeue pointer */
+ trbs_freed = xhci_num_trbs_to(ep_ring->deq_seg, ep_ring->dequeue,
+ td->last_trb_seg, td->last_trb,
+ ep_ring->num_segs);
+ if (trbs_freed < 0)
+ xhci_dbg(xhci, "Failed to count freed trbs at TD finish\n");
+ else
+ ep_ring->num_trbs_free += trbs_freed;
ep_ring->dequeue = td->last_trb;
ep_ring->deq_seg = td->last_trb_seg;
- ep_ring->num_trbs_free += td->num_trbs - 1;
inc_deq(xhci, ep_ring);
return xhci_td_cleanup(xhci, td, ep_ring, td->status);
Patches currently in stable-queue which might be from mathias.nyman@xxxxxxxxxxxxxxx are
queue-6.3/xhci-mem-carefully-calculate-size-for-memory-allocat.patch
queue-6.3/xhci-fix-incorrect-tracking-of-free-space-on-transfer-rings.patch
queue-6.3/xhci-pci-only-run-d3cold-avoidance-quirk-for-s2idle.patch
queue-6.3/xhci-avoid-pci-msi-msix-interrupt-reinitialization-a.patch
