Commit Graph

15716 Commits

Author SHA1 Message Date
Avi Kivity
4aa63af149 pc: grab system_memory
While eventually this should come from the machine initialization function,
take a short cut to avoid converting all machines now.

Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2011-07-29 08:25:44 -05:00
Avi Kivity
bf3fb0e12a ioport: register ranges by byte aligned addresses always
The I/O port space is byte addressable, even for word and long accesses.

An example is the VMware svga card, which has long ports on offsets 0,
1, and 2.

Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2011-07-29 08:25:44 -05:00
Avi Kivity
62152b8a01 exec.c: initialize memory map
Allocate the root memory region and initialize it.

Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2011-07-29 08:25:44 -05:00
Avi Kivity
4ef4db8603 memory: transaction API
Allow changes to the memory hierarchy to be accumulated and
made visible all at once.  This reduces computational effort,
especially when an accelerator (e.g. kvm) is involved.

Useful when a single register update causes multiple changes
to an address space.

Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2011-07-29 08:25:43 -05:00
Avi Kivity
b8af1afbfb memory: separate building the final memory map into two steps
Instead of adding and deleting regions in one pass, do a delete
pass followed by an add pass.  This fixes the following case:

from:
  0x0000-0x0fff ram  (a1)
  0x1000-0x1fff mmio (a2)
  0x2000-0x2fff ram  (a3)

to:
  0x0000-0x2fff ram  (b1)

The single pass algorithm removed a1, added b2, then removed a2 and a3,
which caused the wrong memory map to be built.  The two pass algorithm
removes a1, a2, and a3, then adds b1.

Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2011-07-29 08:25:43 -05:00
Avi Kivity
3e9d69e737 memory: add ioeventfd support
As with the rest of the memory API, the caller associates an eventfd
with an address, and the memory API takes care of registering or
unregistering when the address is made visible or invisible to the
guest.

Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2011-07-29 08:25:43 -05:00
Avi Kivity
74901c3bd0 memory: add backward compatibility for old mmio registration
This eases the transition to the new API.

Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2011-07-29 08:25:43 -05:00
Avi Kivity
627a0e90dc memory: add backward compatibility for old portio registration
Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2011-07-29 08:25:43 -05:00
Avi Kivity
658b222401 memory: I/O address space support
Allow registering I/O ports via the same mechanism as mmio ranges.

Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2011-07-29 08:25:43 -05:00
Avi Kivity
16ef61c9e5 memory: late initialization of ram_addr
For non-RAM memory regions, we cannot tell whether this is an I/O region
or an MMIO region.  Since the qemu backing registration is different for
the two, we have to defer initialization until we know which address
space we are in.

These shenanigans will be removed once the backing registration is unified
with the memory API.

Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2011-07-29 08:25:43 -05:00
Avi Kivity
14a3c10ac8 memory: rename MemoryRegion::has_ram_addr to ::terminates
I/O regions will not have ram_addrs, so this is a better name.

Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2011-07-29 08:25:43 -05:00
Avi Kivity
cc31e6e7b3 memory: abstract address space operations
Prepare for multiple address space support by abstracting away the details
of registering a memory range with qemu's flat representation into an
AddressSpace object.

Note operations which are memory specific are not abstracted, since they will
never be called on I/O address spaces anyway.

Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2011-07-29 08:25:42 -05:00
Avi Kivity
1c0ffa58af Internal interfaces for memory API
get_system_memory() provides the root of the memory hierarchy.

This interface is intended to be private between memory.c and exec.c.
If this file is included elsewhere, it should be regarded as a bug (or
TODO item).  However, it will be temporarily needed for the conversion
to hierarchical memory routing.

Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2011-07-29 08:25:42 -05:00
Avi Kivity
3d8e6bf977 memory: merge adjacent segments of a single memory region
Simple implementations of memory routers, for example the Cirrus VGA memory banks
or the 440FX PAM registers can generate adjacent memory regions which are contiguous.
Detect these and merge them; this saves kvm memory slots and shortens lookup times.

Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2011-07-29 08:25:42 -05:00
Avi Kivity
5a58334769 memory: implement dirty tracking
Currently dirty tracking is implemented by passing through
all calls to the underlying cpu_physical_memory_*() calls.

Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2011-07-29 08:25:42 -05:00
Avi Kivity
093bc2cd88 Hierarchical memory region API
The memory API separates the attributes of a memory region (its size, how
reads or writes are handled, dirty logging, and coalescing) from where it
is mapped and whether it is enabled.  This allows a device to configure
a memory region once, then hand it off to its parent bus to map it according
to the bus configuration.

Hierarchical registration also allows a device to compose a region out of
a number of sub-regions with different properties; for example some may be
RAM while others may be MMIO.

Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2011-07-29 08:25:42 -05:00
Avi Kivity
9d3a4736cb Add memory API documentation
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2011-07-29 08:25:42 -05:00
Jan Kiszka
5ab28c8340 qdev: Reset hot-plugged devices
Device models rely on the core invoking their reset handlers after init.
We do this in the cold-plug case, but so far we miss this step after
hot-plug.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2011-07-29 08:17:40 -05:00
Isaku Yamahata
43e86c8f5b pcie_host: verify mmcfg address range
For a conventional pci device behind
a pcie-to-pci bridge, pci_host handlers get confused by
an out of bounds access in the range [256, 4K).

Check for such an access and make it have no effect.

Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2011-07-29 08:01:52 +03:00
Amit Shah
845f85fa15 virtio: Plug memleak by freeing vdev
virtio_common_init() allocates RAM for the vdev struct (and any
additional memory, depending on the size passed to the function).  This
memory wasn't being freed until now.

Signed-off-by: Amit Shah <amit.shah@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2011-07-27 20:19:27 +03:00
Amit Shah
b52dfd71f3 virtio-net: don't use vdev after virtio_cleanup
virtio_cleanup() will be changed by the following patch to remove the
VirtIONet struct that gets allocated via virtio_common_init().  Ensure
we don't dereference the structure after calling the cleanup function.

Signed-off-by: Amit Shah <amit.shah@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2011-07-27 20:18:47 +03:00
Amit Shah
d92551f28e virtio-blk: Fix memleak on exit
Calling virtio_cleanup() will free up memory allocated in
virtio_common_init().

Signed-off-by: Amit Shah <amit.shah@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2011-07-27 20:08:19 +03:00
Michael S. Tsirkin
1129714ff4 virtio-pci: use generic logic for command access
In practice, guests don't generate config requests
that cross a word boundary, so the logic to
detect command word access is correct because
PCI_COMMAND is 0x4. But depending on this is
tricky, further, it will break with guests
that do try to generate a misaligned access
as we pass it to devices without splitting.

Better to use the generic range_covers_byte for this.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2011-07-27 20:08:15 +03:00
Juan Quintela
ec67464c4f xen_mapcache: remove unused variable
Signed-off-by: Juan Quintela <quintela@redhat.com>
Acked-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
2011-07-27 15:57:09 +01:00
Michael Roth
016c77ad62 Makefile: add missing deps on $(GENERATED_HEADERS)
This fixes a build issue with make -j6+ due to qapi-generated files
being built before $(GENERATED_HEADERS) have been created.

Tested-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com>
Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
2011-07-27 15:57:09 +01:00
Zhi Yong Wu
c20cdf8b91 qmp: fix efect -> effect typo in qmp-commands.hx
Signed-off-by: Zhi Yong Wu <wuzhy@linux.vnet.ibm.com>
Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
2011-07-27 15:57:09 +01:00
Stefan Weil
cf2846b5fa slirp: Fix unusual "comments" in unused code
cppcheck detected two rather strange comments which were not
correctly written as C comments.

They did not cause any harm because they were framed by
#ifdef notdef ... #endif, so they were never compiled.

Fix them nevertheless (we could also remove the unused code).

Signed-off-by: Stefan Weil <weil@mail.berlios.de>
Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
2011-07-27 10:10:17 +01:00
Alexandre Raymond
45b75ae4ee Makefile: Minor cscope fixups
Create cscope symbols for assembly files in addition to .c/.h files.
Create cscope database with full path instead of relative path so cscope
can be used with CSCOPE_DB in any directory.

Signed-off-by: Alexandre Raymond <cerbere@gmail.com>
Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
2011-07-27 10:09:56 +01:00
Wolfgang Mauerer
023367e6cd vhost build fix for i386
vhost.c uses __sync_fetch_and_and(), which is only
available for -march=i486 and above (see
https://bugzilla.redhat.com/show_bug.cgi?id=624279).

Signed-off-by: Wolfgang Mauerer <wolfgang.mauerer@siemens.com>
Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
2011-07-27 08:58:56 +01:00
Jan Kiszka
42e4126b79 pci: Common overflow prevention
Introduce pci_config_read/write_common helpers to prevent passing
accesses down the callback chain that go beyond the config space limits.
Adjust length assertions as they are no longer correct (cutting may
generate valid 3 byte accesses).

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2011-07-27 10:57:22 +03:00
Alon Levy
0f94d6da35 libcacard: add pc file, install it + includes
Additionally:
 + add --includedir configure parameters
 + make install-libcacard install vscclient as well
2011-07-26 10:42:13 +03:00
Stefano Stabellini
679f4f8b17 xen: implement unplug protocol in xen_platform
The unplug protocol is necessary to support PV drivers in the guest: the
drivers expect to be able to "unplug" emulated disks and nics before
initializing the Xen PV interfaces.
It is responsibility of the guest to make sure that the unplug is done
before the emulated devices or the PV interface start to be used.

We use pci_for_each_device to walk the PCI bus, identify the devices and
disks that we want to disable and dynamically unplug them.

Changes in v2:

- use PCI_CLASS constants;

- replace pci_unplug_device with qdev_unplug;

- do not import hw/ide/internal.h in xen_platform.c;

Changes in v3:

- introduce piix3-ide-xen, that support hot-unplug;

- move the unplug code to hw/ide/piix.c;

- just call qdev_unplug from xen_platform.c to unplug the IDE disks;

Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Acked-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Alexander Graf <agraf@suse.de>
2011-07-26 07:02:17 +02:00
Anthony PERARD
834e76ea1c vl.c: Check the asked ram_size later.
As a Xen guest can have more than 2GB of RAM on a 32bit host, we move
the conditions after than we now if we run one Xen or not.

[agraf] separate xen branch from ram_size check

Signed-off-by: Anthony PERARD <anthony.perard@citrix.com>
Signed-off-by: Alexander Graf <agraf@suse.de>
2011-07-26 07:02:10 +02:00
Anthony PERARD
8a369e20e7 xen: Fix the memory registration to reflect of what is done by Xen.
A Xen guest memory is allocated by libxc. But this memory is not
allocated continuously, instead, it leaves the VGA IO memory space not
allocated, same for the MMIO space (at HVM_BELOW_4G_MMIO_START of size
HVM_BELOW_4G_MMIO_LENGTH).

So to reflect that, we do not register the physical memory for this two
holes. But we still keep only one RAMBlock for the all RAM as it is more
easier than have two separate blocks (1 above 4G). Also this prevent QEMU
from use the MMIO space for a ROM.

Signed-off-by: Anthony PERARD <anthony.perard@citrix.com>
Signed-off-by: Alexander Graf <agraf@suse.de>
2011-07-26 06:57:55 +02:00
Anthony PERARD
f15fbc4bd1 cpu-common: Have a ram_addr_t of uint64 with Xen.
In Xen case, memory can be bigger than the host memory. that mean a
32bits host (and QEMU) should be able to handle a RAM address of 64bits.

Signed-off-by: Anthony PERARD <anthony.perard@citrix.com>
Signed-off-by: Alexander Graf <agraf@suse.de>
2011-07-26 06:57:28 +02:00
Alexander Graf
59d21e537b xen: make xen_enabled even more clever
When using xen_enabled() we're currently only checking if xen is enabled
at all during the build. But what if you want to build multiple targets
out of which only one can potentially run xen code?

That means that for generic code we'll still have to fall back to the
variable and potentially slow the code down, but it's not as important as
that is mostly xen device emulation which is not touched for non-xen targets.

The target specific code however can with this patch see that it's unable to
ever execute xen code. We can thus always return 0 on xen_enabled(), giving
gcc enough hints to evict the mapcache code from the target memory management
code.

Signed-off-by: Alexander Graf <agraf@suse.de>
Acked-by: Anthony PERARD <anthony.perard@citrix.com>
2011-07-26 06:43:11 +02:00
Alexander Graf
f03a4ac122 xen: remove CONFIG_XEN_MAPCACHE
We were still exporting CONFIG_XEN_MAPCACHE, even though it's completely
unused by now. Remove it.

Signed-off-by: Alexander Graf <agraf@suse.de>
2011-07-26 06:43:10 +02:00
Anthony PERARD
8ca5692df4 exec.c: Use ram_addr_t in cpu_physical_memory_rw(...).
As the variable pd and addr1 inside the function cpu_physical_memory_rw
are mean to handle a RAM address, they should be of the ram_addr_t type
instead of unsigned long.

Signed-off-by: Anthony PERARD <anthony.perard@citrix.com>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Alexander Graf <agraf@suse.de>
2011-07-26 06:43:10 +02:00
Anthony PERARD
30ab61252b xen: Fix xen_enabled().
Use the "host" CONFIG_ define instead of the "target" one.

Signed-off-by: Anthony PERARD <anthony.perard@citrix.com>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Alexander Graf <agraf@suse.de>
2011-07-26 06:43:10 +02:00
Anthony PERARD
fb4bb2b587 xen: introduce xen_change_state_handler
Remove the call to xenstore_record_dm_state from xen_main_loop_prepare
that is HVM specific.
Add a new vm_change_state_handler shared between xen_pv and xen_hvm
machines to record the VM state to xenstore.

Signed-off-by: Anthony PERARD <anthony.perard@citrix.com>
Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Signed-off-by: Alexander Graf <agraf@suse.de>
2011-07-26 06:43:10 +02:00
Blue Swirl
c886edfb85 Let users select their pythons
Add configure check for python, exit if not found. Add switches
for specifying the path to python, use the path in Makefile.

Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
2011-07-25 16:50:12 +00:00
Blue Swirl
0caf448b80 simpletrace: suppress a warning from unused variable
Avoid this warning:
  CC    simpletrace.o
/src/qemu/simpletrace.c: In function 'writeout_thread':
/src/qemu/simpletrace.c:122:12: error: variable 'unused' set but not used [-Werror=unused-but-set-variable]
by adding GCC attribute unused to the variable.

Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
2011-07-25 14:40:18 +00:00
Blue Swirl
00aa0040e8 Wrap recv to avoid warnings
Avoid warnings like these by wrapping recv():
  CC    slirp/ip_icmp.o
/src/qemu/slirp/ip_icmp.c: In function 'icmp_receive':
/src/qemu/slirp/ip_icmp.c:418:5: error: passing argument 2 of 'recv' from incompatible pointer type [-Werror]
/usr/local/lib/gcc/i686-mingw32msvc/4.6.0/../../../../i686-mingw32msvc/include/winsock2.h:547:32: note: expected 'char *' but argument is of type 'struct icmp *'

Remove also casts used to avoid warnings.

Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
2011-07-25 14:38:56 +00:00
Blue Swirl
aad04cd024 Fix chrdev return value conversion
6e1db57b2a didn't
convert brlapi or win32 chrdevs, breaking build for those.

Fix by converting the chrdevs.

Acked-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
2011-07-24 14:48:57 +00:00
Anthony Liguori
1fc7bd4a86 qemu-ga: remove dependency on gio and gthread
As far as I can tell, there isn't a dependency on gthread.  Also, the only use
of gio was to enable GSocket to accept a unix domain socket.

Since GSocket isn't available on OpenSuSE 11.1, let's just remove that
dependency.

Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2011-07-23 18:24:14 -05:00
Anthony Liguori
4eb36d40da guest-agent: only enable FSFREEZE when it's supported by the kernel
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2011-07-23 18:24:05 -05:00
Anthony Liguori
1167bfd63d Open 1.0 development branch.
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2011-07-23 11:57:53 -05:00
Anthony Liguori
b8095f24f2 Bump version to reflect v0.15.0-rc0
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2011-07-23 11:56:43 -05:00
Matthew Fernandez
8e31bf388e Correct spelling of licensed
Correct typos of "licenced" to "licensed".

Reviewed-by: Stefan Weil <weil@mail.berlios.de>
Reviewed-by: Andreas F=E4rber <andreas.faerber@web.de>
Signed-off-by: Matthew Fernandez <matthew.fernandez@gmail.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2011-07-23 11:26:12 -05:00
Jan Kiszka
d25f89c9e9 Register Linux dyntick timer as per-thread signal
Derived from kvm-tool patch
http://thread.gmane.org/gmane.comp.emulators.kvm.devel/74309

Ingo Molnar pointed out that sending the timer signal to the whole
process, just blocking it everywhere, is suboptimal with an increasing
number of threads. QEMU is also using this pattern so far.

Linux provides a (non-portable) way to restrict the signal to a single
thread: We can use SIGEV_THREAD_ID unless we are forced to emulate
signalfd via an additional thread. That case could theoretically be
optimized as well, but it doesn't look worth bothering.

Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2011-07-23 11:26:12 -05:00