Commit graph

27807 commits

Author SHA1 Message Date
Alexander Graf d02532f08e linux-user: Unlock mmap_lock when resuming guest from page_unprotect
The page_unprotect() function is running everything locked. Before every
potential exit path of the function mmap_unlock() gets called to make sure
we don't leak the lock.

However, the function calls tb_invalidate_phys_page() which again can
exit a signal through longjmp, leaving our mmap_unlock() attempts in vain.

Add a hint to tb_invalidate_phys_page() that we need to unlock before we
can leave back into guest context, so that we don't leak the lock.

This fixes 16-bit i386 wine programs running in linux-user for me.

Signed-off-by: Alexander Graf <agraf@suse.de>
Signed-off-by: Riku Voipio <riku.voipio@linaro.org>
2013-07-23 17:28:28 +03:00
Alexander Graf b24c882b94 linux-user: Reset copied CPUs in cpu_copy() always
When a new thread gets created, we need to reset non arch specific state to
get the new CPU into clean state.

However this reset should happen before the arch specific CPU contents get
copied over. Otherwise we end up having clean reset state in our newly created
thread.

Signed-off-by: Alexander Graf <agraf@suse.de>
Signed-off-by: Riku Voipio <riku.voipio@linaro.org>
2013-07-23 17:28:28 +03:00
Alexander Graf ddaa7e5a2a linux-user: Fix epoll on ARM hosts
The epoll emulation uses data structures without packing them, so the
compiler might choose to add padding inside.

This patch makes the most offending one (target_epoll_event) a packed
structure to make sure we don't pad it by accident. ARM would pad it,
so declare the padding mandatory for ARM targets.

This fixes i386-on-ARM epoll emulation for me.

Signed-off-by: Alexander Graf <agraf@suse.de>
Signed-off-by: Riku Voipio <riku.voipio@linaro.org>
2013-07-23 17:28:28 +03:00
Alexander Graf 732f9e89a1 linux-user: fix segmentation fault passing with h2g(x) != x
When forwarding a segmentation fault into the guest process, we were passing
the host's address directly into the guest process's signal descriptor.

That obviously confused the guest process, since it didn't know what to make
of the (usually 32-bit truncated) address. Passing in h2g(address) makes the
guest process a lot happier.

To make the code more obvious, introduce a h2g_nocheck() macro that does the
same as h2g(), but allows us to convert addresses that may be outside of guest
mapped range into the guest's view of address space.

This fixes java running in arm-linux-user for me.

Signed-off-by: Alexander Graf <agraf@suse.de>
Signed-off-by: Riku Voipio <riku.voipio@linaro.org>
2013-07-23 17:28:28 +03:00
Peter Maydell 82f05b69e6 linux-user: Fix pipe syscall return for SPARC
SPARC is one of the CPUs which has a funny syscall ABI for the
pipe syscall; add it to the set of special cases in do_pipe().

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Riku Voipio <riku.voipio@linaro.org>
2013-07-23 17:28:28 +03:00
Peter Maydell c7819dfbd2 linux-user: Fix target_stat and target_stat64 for OpenRISC
OpenRISC uses the asm-generic versions of target_stat and
target_stat64, but it was incorrectly using the x86/ARM/etc version
due to a misplaced defined(TARGET_OPENRISC).  The previously unused
OpenRISC section of the ifdef ladder also defined an incorrect
target_stat and omitted the target_stat64 definition.  Fix
target_stat, provide target_stat64, and add a comment noting that
these are the asm-generic versions for the benefit of future ports.

Reviewed-by: Jia Liu <proljc@gmail.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Riku Voipio <riku.voipio@linaro.org>
2013-07-23 17:28:28 +03:00
Juan Quintela 9781c3716a migration: add autoconvergence documentation
This hunk got lost during merge.  It is documentation.

Spotted-by: Peter Lieven <lieven-lists@dlhnet.de>
Signed-off-by: Juan Quintela <quintela@redhat.com>
2013-07-23 15:21:09 +02:00
Orit Wasserman 3e47c249b9 Fix real mode guest segments dpl value in savevm
Older KVM version put invalid value in the segments registers dpl field for
real mode guests (0x3).
This breaks migration from those hosts to hosts with unrestricted guest support.
We detect it by checking CS dpl value for real mode guest and fix the dpl values
of all the segment registers.

Signed-off-by: Orit Wasserman <owasserm@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
2013-07-23 14:36:10 +02:00
Orit Wasserman 444ba6797e Fix real mode guest migration
Older KVM versions save CS dpl value to an invalid value for real mode guests
(0x3). This patch detect this situation when loading CPU state and set all the
segments dpl to zero.
This will allow migration from older KVM on host without unrestricted guest
to hosts with restricted guest support.
For example migration from a Penryn host (with kernel 2.6.32) to
a Westmere host (for real mode guest) will fail with "kvm: unhandled exit 80000021".

Signed-off-by: Orit Wasserman <owasserm@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
2013-07-23 14:36:10 +02:00
Michael R. Hines ed4fbd1082 rdma: account for the time spent in MIG_STATE_SETUP through QMP
Using the previous patches, we're now able to timestamp the SETUP
state. Once we have this time, let the user know about it in the
schema.

Reviewed-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Michael R. Hines <mrhines@us.ibm.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
2013-07-23 13:06:37 +02:00
Michael R. Hines 29ae8a4133 rdma: introduce MIG_STATE_NONE and change MIG_STATE_SETUP state transition
As described in the previous patch, until now, the MIG_STATE_SETUP
state was not really a 'formal' state. It has been used as a 'zero' state
(what we're calling 'NONE' here) and QEMU has been unconditionally transitioning
into this state when the QMP migration command was called. Instead we want to
introduce MIG_STATE_NONE, which is our starting state in the state machine, and
then immediately transition into the MIG_STATE_SETUP state when the QMP migrate
command is issued.

In order to do this, we must delay the transition into MIG_STATE_ACTIVE until
later in the migration_thread(). This is done to be able to timestamp the amount of
time spent in the SETUP state for proper accounting to the user during
an RDMA migration.

Furthermore, the management software, until now, has never been aware of the
existence of the SETUP state whatsoever. This must change, because, timing of this
state implies that the state actually exists.

These two patches cannot be separated because the 'query_migrate' QMP
switch statement needs to know how to handle this new state transition.

Reviewed-by: Juan Quintela <quintela@redhat.com>
Tested-by: Michael R. Hines <mrhines@us.ibm.com>
Signed-off-by: Michael R. Hines <mrhines@us.ibm.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
2013-07-23 13:06:37 +02:00
Michael R. Hines d58f574bf3 rdma: allow state transitions between other states besides ACTIVE
This patch is in preparation for the next ones: Until now the MIG_STATE_SETUP
state was not really a 'formal' state. It has been used as a 'zero' state
and QEMU has been unconditionally transitioning into this state when
the QMP migrate command was called. In preparation for timing this state,
we have to make this state a a 'real' state which actually gets transitioned
from later in the migration_thread() from SETUP => ACTIVE, rather than just
automatically dropping into this state at the beginninig of the migration.

This means that the state transition function (migration_finish_set_state())
needs to be capable of transitioning from valid states _other_ than just
MIG_STATE_ACTIVE.

The function is in fact already capable of doing that, but was not allowing the
old state to be a parameter specified as an input.

This patch fixes that and only makes the transition if the current state
matches the old state that the caller intended to transition from.

Reviewed-by: Juan Quintela <quintela@redhat.com>
Tested-by: Michael R. Hines <mrhines@us.ibm.com>
Signed-off-by: Michael R. Hines <mrhines@us.ibm.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
2013-07-23 13:06:28 +02:00
Michael R. Hines 0033b8b42a rdma: send pc.ram
This takes advantages of the previous patches:

1. use the new QEMUFileOps hook 'save_page'

2. call out to the right accessor methods to invoke
   the iteration hooks defined in QEMUFileOps

Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Chegu Vinod <chegu_vinod@hp.com>
Tested-by: Chegu Vinod <chegu_vinod@hp.com>
Tested-by: Michael R. Hines <mrhines@us.ibm.com>
Signed-off-by: Michael R. Hines <mrhines@us.ibm.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
2013-07-23 11:12:00 +02:00
Michael R. Hines 2da776db48 rdma: core logic
Code that does need to be visible is kept
well contained inside this file and this is the only
new additional file to the entire patch.

This file includes the entire protocol and interfaces
required to perform RDMA migration.

Also, the configure and Makefile modifications to link
this file are included.

Full documentation is in docs/rdma.txt

Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Chegu Vinod <chegu_vinod@hp.com>
Tested-by: Chegu Vinod <chegu_vinod@hp.com>
Tested-by: Michael R. Hines <mrhines@us.ibm.com>
Signed-off-by: Michael R. Hines <mrhines@us.ibm.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
2013-07-23 11:12:00 +02:00
Michael R. Hines 44c3b58cf9 rdma: introduce ram_handle_compressed()
This gives RDMA shared access to madvise() on the destination side
when an entire chunk is found to be zero.

Reviewed-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Chegu Vinod <chegu_vinod@hp.com>
Tested-by: Chegu Vinod <chegu_vinod@hp.com>
Tested-by: Michael R. Hines <mrhines@us.ibm.com>
Signed-off-by: Michael R. Hines <mrhines@us.ibm.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
2013-07-23 11:11:59 +02:00
Michael R. Hines de7b685c9e rdma: bugfix: ram_control_save_page()
We were not checking for a valid 'bytes_sent' pointer before accessing it.

Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Michael R. Hines <mrhines@us.ibm.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
2013-07-23 11:11:59 +02:00
Michael R. Hines a5f56b906e rdma: update documentation to reflect new unpin support
As requested, the protocol now includes memory unpinning support.
This has been implemented in a non-optimized manner, in such a way
that one could devise an LRU or other workload-specific information
on top of the basic mechanism to influence the way unpinning happens
during runtime.

The feature is not yet user-facing, and is thus can only be enabled
at compile-time.

Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Michael R. Hines <mrhines@us.ibm.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
2013-07-23 11:11:59 +02:00
Andreas Färber 6f152e9bc8 linux-user: Use X86CPU property to retrieve CPUID family
Avoids duplicating the calculation.

Reviewed-by: Igor Mammedov <imammedo@redhat.com>
Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 02:41:33 +02:00
Andreas Färber 22169d415a gdbstub: Change gdb_register_coprocessor() argument to CPUState
Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 02:41:33 +02:00
Andreas Färber eac8b355f0 cpu: Move gdb_regs field from CPU_COMMON to CPUState
Prepares for changing gdb_register_coprocessor() argument to CPUState.

Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 02:41:33 +02:00
Andreas Färber 2e0f2cfba6 gdbstub: Change GDBState::{c,g}_cpu and find_cpu() to CPUState
Use CPUState::env_ptr where still needed.

Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 02:41:33 +02:00
Andreas Färber f3659eee05 cpu: Introduce CPUClass::memory_rw_debug() for target_memory_rw_debug()
Make inline target_memory_rw_debug() always available and change its
argument to CPUState. Let it check if CPUClass::memory_rw_debug provides
a specialized callback and fall back to cpu_memory_rw_debug() otherwise.

The only overriding implementation is for 32-bit sparc.

This prepares for changing GDBState::g_cpu to CPUState.

Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 02:41:33 +02:00
Andreas Färber f17ec444c3 exec: Change cpu_memory_rw_debug() argument to CPUState
Propagate X86CPU in kvmvapic for simplicity.

Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 02:41:33 +02:00
Andreas Färber 00b941e581 cpu: Turn cpu_get_phys_page_debug() into a CPUClass hook
Change breakpoint_invalidate() argument to CPUState alongside.

Since all targets now assign a softmmu-only field, we can drop helpers
cpu_class_set_{do_unassigned_access,vmsd}() and device_class_set_vmsd().

Prepares for changing cpu_memory_rw_debug() argument to CPUState.

Acked-by: Max Filippov <jcmvbkbc@gmail.com> (for xtensa)
Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 02:41:33 +02:00
Andreas Färber 385b9f0e4d gdbstub: Change gdb_{read,write}_register() argument to CPUState
Use CPUState::env_ptr for now.

Prepares for changing GDBState::g_cpu to CPUState.

Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 02:41:33 +02:00
Andreas Färber db6b81d436 gdbstub: Change gdb_handlesig() argument to CPUState
Prepares for changing GDBState::c_cpu to CPUState.

Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 02:41:33 +02:00
Andreas Färber 9e0c5422cf gdbstub: Change syscall callback argument to CPUState
Callback implementations were specific to arm and m68k, so can easily
cast to ARMCPU and M68kCPU respectively.

Prepares for changing GDBState::c_cpu to CPUState.

Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 02:41:32 +02:00
Andreas Färber 6227881415 kvm: Change kvm_{insert,remove}_breakpoint() argument to CPUState
CPUArchState is no longer directly used since converting CPU loops to
CPUState.

Prepares for changing GDBState::c_cpu to CPUState.

Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 02:41:32 +02:00
Andreas Färber 3825b28ff1 cpu: Change cpu_single_step() argument to CPUState
Use CPUState::env_ptr for now.

Needed for GdbState::c_cpu.

Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 02:41:32 +02:00
Andreas Färber 5ca666c765 gdbstub: Update gdb_handlesig() and gdb_signalled() Coding Style
In particular reindent to 4 instead of 2 spaces.

Prepares for changing cpu_single_step() argument in gdb_handlesig().

Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 02:41:32 +02:00
Andreas Färber ed2803da58 cpu: Move singlestep_enabled field from CPU_COMMON to CPUState
Prepares for changing cpu_single_step() argument to CPUState.

Acked-by: Michael Walle <michael@walle.cc> (for lm32)
Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 02:41:32 +02:00
Richard Henderson 801c4c287b target-alpha: Copy implver to DisasContext
Which allows removing env from DisasContext.

Signed-off-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 02:41:32 +02:00
Richard Henderson ca6862a67f target-alpha: Copy singlestep_enabled to DisasContext
Prepare for removing env from DisasContext.

Signed-off-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 02:41:32 +02:00
Andreas Färber bdf7ae5bbd cpu: Introduce CPUClass::synchronize_from_tb() for cpu_pc_from_tb()
Where no extra implementation is needed, fall back to CPUClass::set_pc().

Acked-by: Michael Walle <michael@walle.cc> (for lm32)
Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 02:41:32 +02:00
Andreas Färber b42eab27be target-unicore32: Implement CPUClass::set_pc()
This adds support for GDB's c addr (Continue) and s addr (Single Step).

Prepares for dropping cpu_pc_from_tb().

Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 02:41:32 +02:00
Andreas Färber a10b978c42 target-moxie: Implement CPUClass::set_pc()
This adds support for GDB's c addr (Continue) and s addr (Single Step).

Prepares for dropping cpu_pc_from_tb().

Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 02:41:32 +02:00
Andreas Färber e700604df0 target-m68k: Implement CPUClass::set_pc()
This adds support for GDB's c addr (Continue) and s addr (Single Step).

Prepares for dropping cpu_pc_from_tb().

Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 02:41:31 +02:00
Andreas Färber f45748f10e cpu: Introduce CPUClass::set_pc() for gdb_set_cpu_pc()
This moves setting the Program Counter from gdbstub into target code.
Use vaddr type as upper-bound replacement for target_ulong.

Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 02:41:31 +02:00
Peter Maydell 2be8d45098 HACKING: Document vaddr type usage
Also extend documentation of target_ulong and abi_ulong.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 02:41:31 +02:00
Andreas Färber 577f42c0e1 cpu: Introduce vaddr type
vaddr is to target_ulong what uintmax_t is to unsigned int.

Its purpose is to allow turning per-target functions with target_ulong
arguments into CPUClass hooks.

Suggested-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 02:41:31 +02:00
Andreas Färber 52f34623b4 gdbstub: Change GDBState::query_cpu to CPUState
Since first_cpu/next_cpu are CPUState, CPUArchState is no longer needed.

This resolves a NULL pointer dereference of query_cpu, introduced
with commit 182735efaf and reported by
TeLeMan and Max Filippov.

Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 02:38:32 +02:00
Hu Tao 52785d9951 isa-bus: Drop isabus_bridge_init() since it does nothing
This works since commit 4ce5dae88e.

Signed-off-by: Hu Tao <hutao@cn.fujitsu.com>
[AF: Edited]
Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 00:37:35 +02:00
Hu Tao f5ba752371 ioapic: Use QOM realize for ioapic
Signed-off-by: Hu Tao <hutao@cn.fujitsu.com>
[AF: Tweaked error message]
Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 00:37:35 +02:00
Hu Tao c118d44b6a kvmvapic: Use QOM realize
Signed-off-by: Hu Tao <hutao@cn.fujitsu.com>
[AF: Renamed variable]
Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 00:37:35 +02:00
Hu Tao 913bc63844 kvm/clock: Use QOM realize for kvmclock
Signed-off-by: Hu Tao <hutao@cn.fujitsu.com>
Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 00:37:35 +02:00
Hu Tao 726887ef44 hpet: Use QOM realize for hpet
Signed-off-by: Hu Tao <hutao@cn.fujitsu.com>
[AF: Avoided SYS_BUS_DEVICE() in loop]
Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 00:37:35 +02:00
Andreas Färber caad4eb345 scsi: Improve error propagation for scsi_bus_legacy_handle_cmdline()
Let scsi_bus_legacy_add_drive() and scsi_bus_legacy_handle_cmdline()
return an Error**. Prepare qdev initfns for QOM realize error model.

Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 00:37:35 +02:00
Andreas Färber 22d6aa03fd megasas: Legacy command line handling fix
Only apply legacy command line handling when the device has not been
hot-plugged. Propagate failure of legacy command line handling.

Cc: qemu-stable@nongnu.org
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 00:37:35 +02:00
Hu Tao b09318cac6 scsi/esp: Use QOM realize for scsi esp
Signed-off-by: Hu Tao <hutao@cn.fujitsu.com>
[AF: Drop sysbus_esp_init()]
Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 00:37:35 +02:00
Hu Tao 563839557c fw_cfg: Use QOM realize for fw_cfg
Signed-off-by: Hu Tao <hutao@cn.fujitsu.com>
[AF: Moved sysbus_init_mmio() to instance_init, renamed variable]
Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-07-23 00:37:35 +02:00