msi depends on pci but pci should not depend on msi.
The only dependency we have is a recent addition
of pci_msi_ functions, IMO they add little enough to
open-code in the small number of users.
Follow-up patches add more cleanups.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
pcie aer needs SERR bit to be writable, and the PCI spec requires
this as well. For compatibility, introduce compat global property
command_serr_enable and make this bit readonly for a pre 0.14 pc
machine.
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Take into account secondary bus reset bit for
bus walk: devices behind a reset bus should not
respond to configuration cycles.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
PCI devices with different vendor or device ids sometimes share
the same rom code. Only the ids and the checksum
differs in a boot rom for such devices.
The i825xx ethernet controller family is a typical example
which is implemented in hw/eepro100.c. It uses at least
3 different device ids, so normally 3 boot roms would be needed.
By automatically patching vendor id and device id (and the checksum)
in qemu, all emulated family members can share the same boot rom.
VGA bios roms are another example with different vendor and device ids.
Only qemu's built-in default rom files will be patched.
v2:
* Patch also the vendor id (and remove the sanity check for vendor id).
v3:
* Don't patch a rom file when its name was set by the user.
Thus we avoid modifications of unknown rom data.
Cc: Gerd Hoffmann <kraxel@redhat.com>
Cc: Markus Armbruster <armbru@redhat.com>
Cc: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Stefan Weil <weil@mail.berlios.de>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Bits 12 to 15 in bridge control register are reserver and must be
read-only zero, curent mask is 0xffff which makes them writeable. Fix
this up by using symbolic bit names for writeable bits instead of a
hardcoded constant.
Fix a comment w1mask -> w1cmask as well.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
This patch adds W1C bit support in the initialization/reset of pci
status registers.
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
This patch fixes hot unplug of cold plugged devices
(those present at system start), which got broken by
5beb8ad503 .
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Cam Macdonell <cam@cs.ualberta.ca>
Tested-by: Cam Macdonell <cam@cs.ualberta.ca>
Reported-by: Cam Macdonell <cam@cs.ualberta.ca>.
pcibus_dev_print() was erroneously retrieving the device bus
number from the secondary bus number offset of the device
instead of the bridge above the device. This ends of landing
in the 2nd byte of the 3rd BAR for devices, which thankfully
is usually zero.
Note: pcibus_get_dev_path() copied this code,
inheriting the same bug. pcibus_get_dev_path() is used for
ramblock naming, so changing it can effect migration. However,
I've only seen this byte be non-zero for an assigned device,
which can't migrate anyway, so hopefully we won't run into
any issues.
This patch does not touch pcibus_get_dev_path, as
bus number is guest assigned for nested buses,
so using it for migration is broken anyway.
Fix it properly later.
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
- save/restore must not check w1c bits
since they are in fact guest controlled
- clear w1c bits on reset
Note: for express there are different kinds of
reset, some leave part of config space alone.
We will likely need a sticky bit mask to implement this.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
use pci_clear_bit_word() in pci_device_reset() where appropriate.
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
this patch implements helper functions to handle msi-x and msi
uniformly.
They will be used later.
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Clear w1cmask when deleting a pci capability.
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Implement RW1C register framework.
With this patch, it would be easy to implement
W1C(Write 1 to Clear) register by just setting w1cmask.
Later RW1C register will be used by pcie.
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Extract range functions from pci.h. These will be used by later patches
by non-PCI devices. Adjust current users.
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
Abort when invalid value for region_num is passed to pci_register_bar.
That is caller's bug. Abort instead of silently ignoring invalid value.
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
This patch sorts out invalid use of pcibus_t.
In pci_register_bar(), pcibus_t wmask is used. It should,
however, be uint64_t because it is used to set
pci configuration space value(PCIDevice::wmask)
by pci_set_quad() or pci_set_long().
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
make pci_parse_devfn() aware of func. With func = NULL it behave as before.
This will be used later.
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
call hotplug callback even when not hotplug case for later use.
And move hotplug check into hotplug callback.
PCIE slot needs this for card presence detection.
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
By making pci_add_capability() the special case of
pci_add_capability_at_offset() of offset = 0,
consolidate pci_add_capability_at_offset() into pci_add_capability().
Cc: Stefan Weil <weil@mail.berlios.de>
Cc: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Move pci bridge related code into pci_bridge.c from pci.c
for further enhancement. pci.c is big enough now, so split it out.
No code change but exporting some accesser functions.
In fact, few pci bridge functions stays in pci.c.
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
move out pci internal structures, PCIBus, PCIBridge and pci_bus_info into
private header file, pci_internals.h.
This is a preparation. Later pci bridge implementation will be
split out form pci.c into pci_bridge.c.
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
make pci hotplug callback return value to caller.
And when returning error, allocated resources are freed.
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Clear interrupt disable bit on reset, according to PCI spec.
Fix pci_device_reset() with 64bit BAR.
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Set PCI multi-function bit according to multifunction property.
PCI address, devfn ,is exported to users as addr property,
so users can populate pci function(PCIDevice in qemu)
at arbitrary devfn.
It means each function(PCIDevice) don't know whether pci device
(PCIDevice[8]) is multi function or not.
So this patch allows user to set multifunction bit via property
and checks whether multifunction bit is set correctly.
Cc: Juan Quintela <quintela@redhat.com>
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
make pci bridge aware of pci multi function property and let pci generic
code to set the bit.
Cc: Blue Swirl <blauwirbel@gmail.com>
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
introduce multifunction property.
Also introduce new convenient device creation function which
will be used later.
For bisectability this patch doesn't do anything, but sets the property
resulting in no functional changes.
Actual changes will be introduced by later patch.
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
replace PCIDeviceInfo::header_type with is_bridge
as suggested by Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
Auto-assigned-address pci function (passing devfn = -1) is always
single function.
This patch adds assert() to guarantee that auto-assigned-address function
is always single function device at function = 0.
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
These will be used to generate unique id strings for ramblocks. The name
field is required, the device pointer is optional as most callers don't
have a device. When there's no device or the device isn't a child of
a bus implementing BusInfo.get_dev_path, the name should be unique for
the platform.
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
When available, we'd like to be able to access the DeviceState
when registering a savevm. For buses with a get_dev_path()
function, this will allow us to create more unique savevm
id strings.
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This works great for PCI since a <segment>:<bus>:<dev>.<fn> uniquely
describes a global address. No need to traverse up the qdev tree.
PCI segment support is a placeholder for compatibility once we
support multiple segments.
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Setting the ID in pci_nic_init() is a blatant violation of the
DeviceState abstraction. Which even carries a comment advising
against this:
/* This structure should not be accessed directly. We declare it here
so that it can be embedded in individual device state structures. */
What's worse, it bypasses the code ensuring unique qdev IDs: "-device
virtio-net-pci,id=foo -net nic,id=foo -net nic,name=foo" happily
creates three qdevs with ID "foo". That's because qdev relies on
qemu_opts_create() to ensure unique IDs, but -net nic uses a different
QemuOptsList, which means id is in a different namespace. And its
name is not checked for uniqueness at all.
-net nic and pci_add are legacy. Use -device and device_add if you
want a NIC with a qdev ID.
This reverts what's still left of commit eb54b6dc "qdev: add id=
support for pci nics."
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
When mistakenly configuring two devices in the same PCI slot,
QEMU gives a not entirely obvious message about a 'devfn' being
in use:
$ qemu -device rtl8139 -device virtio-balloon-pci,bus=pci.0,addr=0x3
qemu-kvm: -device virtio-balloon-pci,bus=pci.0,addr=0x3: PCI: devfn 24 not available for virtio-balloon-pci, in use by rtl8139
The user does not configure 'devfn' numbers, they use slot+function.
Thus the error messages should be reported back to the user with that
same terminology rather than the internal QEMU terminology. This
patch makes it report:
$ qemu -device rtl8139 -device virtio-balloon-pci,bus=pci.0,addr=0x3
qemu: -device virtio-balloon-pci,bus=pci.0,addr=0x3.7: PCI: slot 3 function 0 not available for virtio-balloon-pci, in use by rtl8139
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Previous commit added QMP documentation to the qemu-monitor.hx
file, it's is a copy of this information.
While it's good to keep it near code, maintaining two copies of
the same information is too hard and has little benefit as we
don't expect client writers to consult the code to find how to
use a QMP command.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
add helper function which converts root bus to pci domain.
make them aware of pci domain.
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Use pci accessor function.
don't return value because it always return 0 and
the caller doesn't check the return value.
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Add argument, DeviceState*, to pci hot plug callback.
The argument will be used later to remove global variable.
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
Code for saving irq_state got vm_state
macros wrong, passing in the wrong parameter.
As a result, we both saved a wrong value
and restored it to a wrong offset.
This leads to device and bus irq counts getting
out of sync, which in turn leads to interrupts getting lost or
never cleared, such as
https://bugzilla.redhat.com/show_bug.cgi?id=588133
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Juan Quintela <quintela@redhat.com>
If the init function of a device fails, as might happen with device
assignment, we never undo the work done by do_pci_register_device().
This not only causes a bit of a memory leak, but also leaves a bogus
pointer in the bus devices array that can cause a segfault or
garbage data from 'info pci'.
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
When looking down child bus, it should look parent bridge's
bus number, not child bus's.
Optimized tail recursion and style fix.
Cc: Blue Swirl <blauwirbel@gmail.com>
Cc: "Michael S. Tsirkin" <mst@redhat.com>
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
