From: Michael Roth <mdroth@linux.vnet.ibm.com>
To: qemu-devel@nongnu.org
Cc: qemu-stable@nongnu.org, "Michael S. Tsirkin" <mst@redhat.com>,
Benjamin Herrenschmidt <benh@kernel.crashing.org>
Subject: [Qemu-devel] [PATCH 53/56] pcie: fix link active status bit migration
Date: Mon, 8 Aug 2016 16:04:24 -0500 [thread overview]
Message-ID: <1470690267-31454-54-git-send-email-mdroth@linux.vnet.ibm.com> (raw)
In-Reply-To: <1470690267-31454-1-git-send-email-mdroth@linux.vnet.ibm.com>
From: "Michael S. Tsirkin" <mst@redhat.com>
We changed link status register in pci express endpoint capability
over time. Specifically,
commit b2101eae63ea57b571cee4a9075a4287d24ba4a4 ("pcie: Set the "link
active" in the link status register") set data link layer link active
bit in this register without adding compatibility to old machine types.
When migrating from qemu 2.3 and older this affects xhci devices which
under machine type 2.0 and older have a pci express endpoint capability
even if they are on a pci bus.
Add compatibility flags to make this bit value match what it was under
2.3.
Additionally, to avoid breaking migration from qemu 2.3 and up,
suppress checking link status during migration: this seems sane
since hardware can change link status at any time.
https://bugzilla.redhat.com/show_bug.cgi?id=1352860
Reported-by: Gerd Hoffmann <kraxel@redhat.com>
Fixes: b2101eae63ea57b571cee4a9075a4287d24ba4a4
("pcie: Set the "link active" in the link status register")
Cc: qemu-stable@nongnu.org
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
(cherry picked from commit 6b4495401bdf442457b713b7e3994b465c55af35)
Conflicts:
hw/pci/pcie.c
* removed functional dependency on 6383292
Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com>
---
hw/pci/pci.c | 2 ++
hw/pci/pcie.c | 15 ++++++++++++++-
include/hw/compat.h | 4 ++++
include/hw/pci/pci.h | 3 +++
4 files changed, 23 insertions(+), 1 deletion(-)
diff --git a/hw/pci/pci.c b/hw/pci/pci.c
index bb605ef..616f04c 100644
--- a/hw/pci/pci.c
+++ b/hw/pci/pci.c
@@ -62,6 +62,8 @@ static Property pci_props[] = {
QEMU_PCI_CAP_MULTIFUNCTION_BITNR, false),
DEFINE_PROP_BIT("command_serr_enable", PCIDevice, cap_present,
QEMU_PCI_CAP_SERR_BITNR, true),
+ DEFINE_PROP_BIT("x-pcie-lnksta-dllla", PCIDevice, cap_present,
+ QEMU_PCIE_LNKSTA_DLLLA_BITNR, true),
DEFINE_PROP_END_OF_LIST()
};
diff --git a/hw/pci/pcie.c b/hw/pci/pcie.c
index 728386a..c85b4f7 100644
--- a/hw/pci/pcie.c
+++ b/hw/pci/pcie.c
@@ -47,6 +47,7 @@ int pcie_cap_init(PCIDevice *dev, uint8_t offset, uint8_t type, uint8_t port)
{
int pos;
uint8_t *exp_cap;
+ uint8_t *cmask;
assert(pci_is_express(dev));
@@ -57,6 +58,7 @@ int pcie_cap_init(PCIDevice *dev, uint8_t offset, uint8_t type, uint8_t port)
}
dev->exp.exp_cap = pos;
exp_cap = dev->config + pos;
+ cmask = dev->cmask + pos;
/* capability register
interrupt message number defaults to 0 */
@@ -80,7 +82,18 @@ int pcie_cap_init(PCIDevice *dev, uint8_t offset, uint8_t type, uint8_t port)
PCI_EXP_LNK_LS_25);
pci_set_word(exp_cap + PCI_EXP_LNKSTA,
- PCI_EXP_LNK_MLW_1 | PCI_EXP_LNK_LS_25 |PCI_EXP_LNKSTA_DLLLA);
+ PCI_EXP_LNK_MLW_1 | PCI_EXP_LNK_LS_25);
+
+ if (dev->cap_present & QEMU_PCIE_LNKSTA_DLLLA) {
+ pci_word_test_and_set_mask(exp_cap + PCI_EXP_LNKSTA,
+ PCI_EXP_LNKSTA_DLLLA);
+ }
+
+ /* We changed link status bits over time, and changing them across
+ * migrations is generally fine as hardware changes them too.
+ * Let's not bother checking.
+ */
+ pci_set_word(cmask + PCI_EXP_LNKSTA, 0);
pci_set_long(exp_cap + PCI_EXP_DEVCAP2,
PCI_EXP_DEVCAP2_EFF | PCI_EXP_DEVCAP2_EETLPP);
diff --git a/include/hw/compat.h b/include/hw/compat.h
index a5dbbf8..81fc19b 100644
--- a/include/hw/compat.h
+++ b/include/hw/compat.h
@@ -73,6 +73,10 @@
.driver = "virtio-rng-pci",\
.property = "any_layout",\
.value = "off",\
+ },{\
+ .driver = TYPE_PCI_DEVICE,\
+ .property = "x-pcie-lnksta-dllla",\
+ .value = "off",\
},
#define HW_COMPAT_2_2 \
diff --git a/include/hw/pci/pci.h b/include/hw/pci/pci.h
index ef6ba51..e7f2df5 100644
--- a/include/hw/pci/pci.h
+++ b/include/hw/pci/pci.h
@@ -173,6 +173,9 @@ enum {
/* PCI Express capability - Power Controller Present */
#define QEMU_PCIE_SLTCAP_PCP_BITNR 7
QEMU_PCIE_SLTCAP_PCP = (1 << QEMU_PCIE_SLTCAP_PCP_BITNR),
+ /* Link active status in endpoint capability is always set */
+#define QEMU_PCIE_LNKSTA_DLLLA_BITNR 8
+ QEMU_PCIE_LNKSTA_DLLLA = (1 << QEMU_PCIE_LNKSTA_DLLLA_BITNR),
};
#define TYPE_PCI_DEVICE "pci-device"
--
1.9.1
next prev parent reply other threads:[~2016-08-08 21:06 UTC|newest]
Thread overview: 62+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-08-08 21:03 [Qemu-devel] [PATCH 00/56] Patch Round-up for stable 2.6.1, freeze on 2016-08-12 Michael Roth
2016-08-08 21:03 ` [Qemu-devel] [PATCH 01/56] i386: kvmvapic: initialise imm32 variable Michael Roth
2016-08-08 21:03 ` [Qemu-devel] [PATCH 02/56] spice/gl: add & use qemu_spice_gl_monitor_config Michael Roth
2016-08-08 21:03 ` [Qemu-devel] [PATCH 03/56] vl: change runstate only if new state is different from current state Michael Roth
2016-08-08 21:03 ` [Qemu-devel] [PATCH 04/56] tools: kvm_stat: Powerpc related fixes Michael Roth
2016-08-08 21:03 ` [Qemu-devel] [PATCH 05/56] exec.c: Ensure right alignment also for file backed ram Michael Roth
2016-08-08 21:03 ` [Qemu-devel] [PATCH 06/56] usb:xhci: no DMA on HC reset Michael Roth
2016-08-08 21:03 ` [Qemu-devel] [PATCH 07/56] target-mips: fix call to memset in soft reset code Michael Roth
2016-08-08 21:03 ` [Qemu-devel] [PATCH 08/56] target-i386: key sfence availability on CPUID_SSE, not CPUID_SSE2 Michael Roth
2016-08-08 21:03 ` [Qemu-devel] [PATCH 09/56] configure: Allow builds with extra warnings Michael Roth
2016-08-08 21:03 ` [Qemu-devel] [PATCH 10/56] migration: regain control of images when migration fails to complete Michael Roth
2016-08-08 21:03 ` [Qemu-devel] [PATCH 11/56] json-streamer: Don't leak tokens on incomplete parse Michael Roth
2016-08-08 21:03 ` [Qemu-devel] [PATCH 12/56] json-streamer: fix double-free on exiting during a parse Michael Roth
2016-08-08 21:03 ` [Qemu-devel] [PATCH 13/56] esp: check command buffer length before write(CVE-2016-4439) Michael Roth
2016-08-08 21:03 ` [Qemu-devel] [PATCH 14/56] esp: check dma length before reading scsi command(CVE-2016-4441) Michael Roth
2016-08-08 21:03 ` [Qemu-devel] [PATCH 15/56] block/nfs: refuse readahead if cache.direct is on Michael Roth
2016-08-08 21:03 ` [Qemu-devel] [PATCH 16/56] usb/ohci: Fix crash with when specifying too many num-ports Michael Roth
2016-08-08 21:03 ` [Qemu-devel] [PATCH 17/56] vga: add sr_vbe register set Michael Roth
2016-08-08 21:03 ` [Qemu-devel] [PATCH 18/56] vfio: Fix broken EEH Michael Roth
2016-08-08 21:03 ` [Qemu-devel] [PATCH 19/56] block/iscsi: avoid potential overflow of acb->task->cdb Michael Roth
2016-08-08 21:03 ` [Qemu-devel] [PATCH 20/56] nbd: Don't trim unrequested bytes Michael Roth
2016-08-08 21:03 ` [Qemu-devel] [PATCH 21/56] savevm: fail if migration blockers are present Michael Roth
2016-08-08 21:03 ` [Qemu-devel] [PATCH 22/56] Fix configure test for PBKDF2 in nettle Michael Roth
2016-08-08 21:03 ` [Qemu-devel] [PATCH 23/56] scsi: pvscsi: check command descriptor ring buffer size (CVE-2016-4952) Michael Roth
2016-08-08 21:03 ` [Qemu-devel] [PATCH 24/56] scsi: mptsas: infinite loop while fetching requests Michael Roth
2016-08-08 21:03 ` [Qemu-devel] [PATCH 25/56] block: Drop bdrv_ioctl_bh_cb Michael Roth
2016-08-08 21:03 ` [Qemu-devel] [PATCH 26/56] vmsvga: move fifo sanity checks to vmsvga_fifo_length Michael Roth
2016-08-08 21:03 ` [Qemu-devel] [PATCH 27/56] vmsvga: add more fifo checks Michael Roth
2016-08-08 21:03 ` [Qemu-devel] [PATCH 28/56] vmsvga: shadow fifo registers Michael Roth
2016-08-08 21:04 ` [Qemu-devel] [PATCH 29/56] vmsvga: don't process more than 1024 fifo commands at once Michael Roth
2016-08-08 21:04 ` [Qemu-devel] [PATCH 30/56] io: remove mistaken call to object_ref on QTask Michael Roth
2016-08-08 21:04 ` [Qemu-devel] [PATCH 31/56] ui: fix regression in printing VNC host/port on startup Michael Roth
2016-08-08 21:04 ` [Qemu-devel] [PATCH 32/56] net: fix qemu_announce_self not emitting packets Michael Roth
2016-08-08 21:04 ` [Qemu-devel] [PATCH 33/56] backup: Don't leak BackupBlockJob in error path Michael Roth
2016-08-08 21:04 ` [Qemu-devel] [PATCH 34/56] qcow2: Avoid making the L1 table too big Michael Roth
2016-08-08 21:04 ` [Qemu-devel] [PATCH 35/56] qapi: Fix crash on missing alternate member of QAPI struct Michael Roth
2016-08-08 21:04 ` [Qemu-devel] [PATCH 36/56] pci-assign: Move "Invalid ROM" error message to pci-assign-load-rom.c Michael Roth
2016-08-08 21:04 ` [Qemu-devel] [PATCH 37/56] vfio/pci: Fix VGA quirks Michael Roth
2016-08-08 21:04 ` [Qemu-devel] [PATCH 38/56] nbd: Allow larger requests Michael Roth
2016-08-08 21:04 ` [Qemu-devel] [PATCH 39/56] scsi-generic: Merge block max xfer len in INQUIRY response Michael Roth
2016-08-08 21:04 ` [Qemu-devel] [PATCH 40/56] scsi: Advertise limits by blocksize, not 512 Michael Roth
2016-08-08 21:04 ` [Qemu-devel] [PATCH 41/56] target-sparc: fix register corruption in ldstub if there is no write permission Michael Roth
2016-08-08 21:04 ` [Qemu-devel] [PATCH 42/56] virtio: set low features early on load Michael Roth
2016-08-08 21:04 ` [Qemu-devel] [PATCH 43/56] Revert "virtio-net: unbreak self announcement and guest offloads after migration" Michael Roth
2016-08-08 21:04 ` [Qemu-devel] [PATCH 44/56] s390x/ipl: fix reboots for migration from different bios Michael Roth
2016-08-08 21:04 ` [Qemu-devel] [PATCH 45/56] blockdev: Fix regression with the default naming of throttling groups Michael Roth
2016-08-08 21:04 ` [Qemu-devel] [PATCH 46/56] qemu-iotests: Test " Michael Roth
2016-08-08 21:04 ` [Qemu-devel] [PATCH 47/56] util: Fix MIN_NON_ZERO Michael Roth
2016-08-08 21:04 ` [Qemu-devel] [PATCH 48/56] block/iscsi: fix rounding in iscsi_allocationmap_set Michael Roth
2016-08-08 21:04 ` [Qemu-devel] [PATCH 49/56] Fix some typos found by codespell Michael Roth
2016-08-08 21:04 ` [Qemu-devel] [PATCH 50/56] nbd: More debug typo fixes, use correct formats Michael Roth
2016-08-08 21:04 ` [Qemu-devel] [PATCH 51/56] nbd: Don't use *_to_cpup() functions Michael Roth
2016-08-08 21:04 ` [Qemu-devel] [PATCH 52/56] nbd: Limit nbdflags to 16 bits Michael Roth
2016-08-08 21:04 ` Michael Roth [this message]
2016-08-08 21:04 ` [Qemu-devel] [PATCH 54/56] target-i386: fix typo in xsetbv implementation Michael Roth
2016-08-08 21:04 ` [Qemu-devel] [PATCH 55/56] virtio: error out if guest exceeds virtqueue size Michael Roth
2016-08-08 21:04 ` [Qemu-devel] [PATCH 56/56] ide: fix halted IO segfault at reset Michael Roth
2016-08-09 18:34 ` John Snow
2016-08-08 23:40 ` [Qemu-devel] [Qemu-stable] [PATCH 00/56] Patch Round-up for stable 2.6.1, freeze on 2016-08-12 Cole Robinson
2016-08-09 20:04 ` Michael Roth
2016-08-13 1:43 ` Gonglei
2016-08-09 20:12 ` [Qemu-devel] " Bruce Rogers
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1470690267-31454-54-git-send-email-mdroth@linux.vnet.ibm.com \
--to=mdroth@linux.vnet.ibm.com \
--cc=benh@kernel.crashing.org \
--cc=mst@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=qemu-stable@nongnu.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).