render / rpms / qemu

Forked from rpms/qemu 10 months ago
Clone
Blob Blame History Raw
From d2901e4798cb106d5b265aa4e3ae05c06bf2bd1c Mon Sep 17 00:00:00 2001
From: Hans de Goede <hdegoede@redhat.com>
Date: Thu, 6 Sep 2012 17:10:48 +0200
Subject: [PATCH 360/366] ehci: Don't process too much frames in 1 timer tick

The Linux ehci isoc scheduling code fills the entire schedule ahead of
time minus 80 frames. If we make a large jump in where we are in the
schedule, ie 40 frames, then the scheduler all of a sudden will only have
40 frames left to work in, causing it to fail packet submissions
with error -27 (-EFBIG).

Note at first I had MAX_FR_PER_TICK set to 8, which works well with newer
Linux guest kernels, but not with older ones (such as the RHEL-6 kernel).

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
---
 hw/usb/hcd-ehci.c | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/hw/usb/hcd-ehci.c b/hw/usb/hcd-ehci.c
index 30d2b56..5398544 100644
--- a/hw/usb/hcd-ehci.c
+++ b/hw/usb/hcd-ehci.c
@@ -140,6 +140,7 @@
 #define NB_PORTS         6        // Number of downstream ports
 #define BUFF_SIZE        5*4096   // Max bytes to transfer per transaction
 #define MAX_QH           100      // Max allowable queue heads in a chain
+#define MAX_FR_PER_TICK  4        /* Max frames to process in one timer tick */
 
 /*  Internal periodic / asynchronous schedule state machine states
  */
@@ -2460,6 +2461,14 @@ static void ehci_frame_timer(void *opaque)
             DPRINTF("WARNING - EHCI skipped %d frames\n", skipped_frames);
         }
 
+        /*
+         * Processing too much frames at once causes the Linux EHCI isoc
+         * scheduling code to fail packet re-submissions with -EFBIG.
+         */
+        if (frames > MAX_FR_PER_TICK) {
+            frames = MAX_FR_PER_TICK;
+        }
+
         for (i = 0; i < frames; i++) {
             ehci_update_frindex(ehci, 1);
             ehci_advance_periodic_state(ehci);
-- 
1.7.12