In KVM mode the bootrom is loaded and executed from the last 1MB of
DRAM.
Based on "[PATCH 12/12] KVM/MIPS: General KVM support and support for
SMP Guests" by Sanjay Lal <sanjayl@kymasys.com>.
Signed-off-by: James Hogan <james.hogan@imgtec.com>
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Cc: Peter Maydell <peter.maydell@linaro.org>
Cc: Sanjay Lal <sanjayl@kymasys.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
COP0 emulation is in-kernel for KVM, so inject IRQ2 (I/O) interrupts via
ioctls.
Signed-off-by: Sanjay Lal <sanjayl@kymasys.com>
Signed-off-by: James Hogan <james.hogan@imgtec.com>
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Reviewed-by: Andreas Färber <afaerber@suse.de>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
When KVM is enabled call kvm_mips_reset_vcpu() from mips_cpu_reset() as
done for other targets since commit 50a2c6e55f (kvm: reset state from
the CPU's reset method).
Signed-off-by: James Hogan <james.hogan@imgtec.com>
Cc: Aurelien Jarno <aurelien@aurel32.net>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Implement the main KVM arch API for MIPS.
Signed-off-by: Sanjay Lal <sanjayl@kymasys.com>
Signed-off-by: James Hogan <james.hogan@imgtec.com>
Cc: Aurelien Jarno <aurelien@aurel32.net>
Cc: Gleb Natapov <gleb@redhat.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Andreas Färber <afaerber@suse.de>
Cc: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
MIPS/Linux is unusual in having 128 signals rather than just 64 like
most other architectures. This means its sigmask is 16 bytes instead of
8, so allow arches to override the sigmask->len value passed to the
KVM_SET_SIGNAL_MASK ioctl in kvm_set_signal_mask() by calling
kvm_set_sigmask_len() from kvm_arch_init(). Otherwise default to 8
bytes.
Signed-off-by: James Hogan <james.hogan@imgtec.com>
Cc: Aurelien Jarno <aurelien@aurel32.net>
Cc: Sanjay Lal <sanjayl@kymasys.com>
Cc: Gleb Natapov <gleb@redhat.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
MIPS KVM trap & emulate mode (which is currently the only supported
mode) has to add an extra kseg0/kseg1 at 0x40000000 and an extra
kseg2/kseg3 at 0x60000000. Take this into account in
get_physical_address() so that debug memory access works.
This is done by translating the address to a standard kseg0 or kseg2
address before doing the normal address translation. The real virtual
address is still used for TLB lookups.
Signed-off-by: James Hogan <james.hogan@imgtec.com>
Cc: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Add preprocessor definitions for 32bit segment bases for use in
get_physical_address(). These will also be taken advantage of in the
next patch which adds KVM awareness.
Signed-off-by: James Hogan <james.hogan@imgtec.com>
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Add API for converting physical addresses to KVM guest KSEG0 addresses,
and fix the existing API for converting KSEG0 addresses to physical
addresses to work in the KVM case. Both have the same sized KSEG0, so
it's just a case of fixing the mask.
In KVM trap and emulate mode both the guest kernel and guest userspace
execute in useg:
Guest User address space: 0x00000000..0x3fffffff
Guest Kernel Unmapped: 0x40000000..0x5fffffff
Guest Kernel Mapped: 0x60000000..0x7fffffff
Signed-off-by: Sanjay Lal <sanjayl@kymasys.com>
Signed-off-by: James Hogan <james.hogan@imgtec.com>
Cc: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Compare/Count timer interrupts are handled in-kernel for KVM. Therefore
don't bother creating the timer at init time if KVM is enabled. This
will conveniently avoid attempts to set the timeout when
cpu_mips_store_count() is called at reset with KVM enabled, treating the
timer as stopped so that CP0_Count is modified directly.
Signed-off-by: Sanjay Lal <sanjayl@kymasys.com>
[james.hogan@imgtec.com: Update after "target-mips: Reset CPU timer
consistently" which moves timer start to reset time]
Signed-off-by: James Hogan <james.hogan@imgtec.com>
Cc: Aurelien Jarno <aurelien@aurel32.net>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The MIPS CPU timer (CP0 Count/Compare registers & QEMU timer) is
reset at machine initialisation, including starting the timeout. Both
registers however are placed before mvp in CPUMIPSState so they will
both be zeroed on reset by the memset in mips_cpu_reset() including soon
after init. This doesn't take into account that the timer may be
running, in which case env->CP0_Count will represent the delta against
the VM clock and the timeout will need updating.
At init time (cpu_mips_clock_init()), lets only create the timer.
Setting Count = 1 and starting the timer (cpu_mips_store_count()) can be
done at reset time from cpu_state_reset(), which is after the memset.
There is also no need to set CP0_Compare = 0 as that is already handled
by the memset.
Note that a reset occurs from mips_cpu_realizefn() which is before the
machine init callback has had a chance to set up the CPU interrupts and
the CPU timer, so env->timer will be NULL. This case is handled
explicitly in cpu_mips_store_count(), treating the timer as disabled
(which will also be the right thing to do when KVM support is added).
Reported-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: James Hogan <james.hogan@imgtec.com>
Cc: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
KVM tells us the number of GSIs it can handle inside the kernel. That value is
basically KVM_MAX_IRQ_ROUTES. However when we try to set the GSI mapping table,
it checks for
r = -EINVAL;
if (routing.nr >= KVM_MAX_IRQ_ROUTES)
goto out;
erroring out even when we're only using all of the GSIs. To make sure we never
hit that limit, let's reduce the number of GSIs we get from KVM by one.
Cc: qemu-stable@nongnu.org
Signed-off-by: Alexander Graf <agraf@suse.de>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
* remotes/bonzini/memory:
memory: Don't call memory_region_update_coalesced_range if nothing changed
memory: MemoryRegion: rename parent to container
memory: MemoryRegion: factor out memory region re-adder
memory: MemoryRegion: factor out subregion add functionality
qtest: fix qtest_clock_warp() for no deadline case
exec: dummy_section: Pass address space through.
memory: Simplify mr_add_subregion() if-else
memory: Don't update all memory region when ioeventfd changed
unset RAMBlock idstr when unregister MemoryRegion
exec: introduce qemu_ram_unset_idstr() to unset RAMBlock idstr
MAINTAINERS: Add myself as Memory API maintainer
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
With huge number of PCI devices in the system (for example, 200
virtio-blk-pci), this unconditional call can slow down emulation of
irrelevant PCI operations drastically, such as a BAR update on a device
that has no coalescing region. So avoid it.
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Store the request and response headers by value, and let
virtio_scsi_parse_req check that there is only one of datain
and dataout.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Retrieve sense and copy it to guest memory, to prepare for when we will use
qemu_iovec_from_buf.
Swap response and request, since we'll use the tail of VirtIOSCSIReq
for the CDB.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
- Introduce virtio_scsi_init_req and virtio_scsi_free_req
- rename qemu_sgl_init_external to qemu_sgl_concat
- move virtio_scsi_parse_req from virtio_scsi_pop_req to callers
and add header length checks to virtio_scsi_parse_req.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
This makes scsi_command_name() public.
This makes use of scsi_command_name() in debug output for scsi-disk and
spapr-vscsi host bus adapter. Before this, SCSI used to print hex numbers
instead of human-friendly strings.
This adds GET_EVENT_STATUS_NOTIFICATION and READ_DISC_INFORMATION to
the list of SCSI commands supported by scsi_command_name().
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
This patch fixes a bug in scsi_block_new_request() that was introduced
by commit 137745c5c6. If the host cache
is used - i.e. if BDRV_O_NOCACHE is _not_ set - the 'break' statement
needs to be executed to 'fall back' to SG_IO.
Cc: qemu-stable@nongnu.org
Signed-off-by: Ulrich Obergfell <uobergfe@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
In scsi-disk.c, if you #define DEBUG_SCSI=1, you get:
hw/scsi/scsi-disk.c: In function 'scsi_disk_emulate_command':
hw/scsi/scsi-disk.c:2018: error: 'SCSIRequest' has no member named 'buf'
Change the debugging statement to match the actual value tested.
Signed-off-by: Paul Janzen <pcj@pauljanzen.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
this patch changes the driver to uses 16 Byte CDBs for
READ/WRITE only if the target requires 64bit lba addressing.
On one hand this saves 6 bytes in each PDU on the other
hand it seems that 10 Byte CDBs seems to be much better
supported and tested as a recent issue I had with a
major storage supplier lined out.
For WRITESAME the logic is a bit more tricky as WRITESAME10
with UNMAP was added really late. Thus a fallback to WRITESAME16
is possible if it supports UNMAP and WRITESAME10 not.
Signed-off-by: Peter Lieven <pl@kamp.de>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
it might happen in the future that a function directly invokes its callback.
In this case we end up in a segfault because the iTask is gone when the BH
is scheduled.
Signed-off-by: Peter Lieven <pl@kamp.de>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
this patch adds handling of BUSY status reponse from an iSCSI target.
Currently, we fail with -EIO in case of SCSI_STATUS_BUSY while the
obvious reaction would be to retry the operation after some time.
The retry time is randomly choosen from a range with exponential
growth increasing with each retry.
This patch includes most of the changes by a an upcoming patch
from Stefan Hajnoczi:
iscsi: implement .bdrv_detach/attach_aio_context()
because I also need the reference to the aio_context for
the retry timer to work. I included the changes to maintain
better mergeability.
Signed-off-by: Peter Lieven <pl@kamp.de>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
memory_region_set_address is mostly just a function that deletes and
re-adds a memory region. Factor this generic functionality out into a
re-usable function. This prepares support for further QOMification
of MemoryRegion.
Signed-off-by: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Split off the core looping code that actually adds subregions into
it's own fn. This prepares support for Memory Region qomification
where setting the MR address or parent via QOM will back onto this more
minimal function.
Signed-off-by: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
[Rename new function. - Paolo]
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Use dedicated qemu_soonest_timeout() instead of MIN().
Signed-off-by: Sergey Fedorov <serge.fdrv@gmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Rather than use the global singleton.
Signed-off-by: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
This if else is not needed. The previous call to memory_region_add
(whether _overlap or not) will always set priority and may_overlap
to desired values. And its not possible to get here without having
called memory_region_add_subregion due to the null guard on parent.
So we can just directly call memory_region_add_subregion_common.
Signed-off-by: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
memory mappings don't rely on ioeventfds, there is no need
to destroy and rebuild them when manipulating ioeventfds,
otherwise it scarifies performance.
according to testing result, each ioeventfd deleing needs
about 5ms, within which memory mapping rebuilding needs
about 4ms. With many Nics and vmchannel in a VM doing migrating,
there can be many ioeventfds deleting which increasing
downtime remarkably.
Signed-off-by: Gonglei <arei.gonglei@huawei.com>
Signed-off-by: Herongguang <herongguang.he@huawei.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
I'm not including Avi since he has already removed himself from the
KVM entry. I'm not going to commit my patches without review.
Acked-by: Edgar E. Iglesias <edgar.iglesias@gmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
In user mode Linux, Qemu currently refuses to load ELF files that do not
contain section headers (ehdr->e_shentsize == 0). Since section headers are not
required in order to load an ELF file, simply removing the e_shentsize check in
elf_check_ehdr() allows ELF binaries with no section headers to be run properly
in user mode:
Signed-off-by: Craig Heffner <cheffner@tacnetsol.com>
Signed-off-by: Riku Voipio <riku.voipio@linaro.org>
This fixes "Cannot open audit interface - aborting." when the
EAFNOSUPPORT errno differs between the target and host
architectures (e.g. mips target and x86_64 host).
Signed-off-by: Ed Swierk <eswierk@skyportsystems.com>
Signed-off-by: Riku Voipio <riku.voipio@linaro.org>
If the guest's "long" type is smaller than the host's, then
our sched_getaffinity wrapper needs to round the buffer size
up to a multiple of the host sizeof(long). This means that when
we copy the data back from the host buffer to the guest's
buffer there might be more than we can fit. Rather than
overflowing the guest's buffer, handle this case by returning
EINVAL or ignoring the unused extra space, as appropriate.
Note that only guests using the syscall interface directly might
run into this bug -- the glibc wrappers around it will always
use a buffer whose size is a multiple of 8 regardless of guest
architecture.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Riku Voipio <riku.voipio@linaro.org>
We were returning the incorrect uname string (with a hyphen, not
an underscore) for x86_64. Fix this by removing the x86_64 special
case, since the default "just use UNAME_MACHINE" behaviour suffices.
This leaves cpu_to_uname_machine() special cases for only those
architectures which need to vary the string based on runtime CPU
features.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Riku Voipio <riku.voipio@linaro.org>
gcc-4.9 finds unused operand:
linux-user/syscall.c: In function ‘host_to_target_stat64’:
linux-user/qemu.h:301:19: error: right-hand operand of comma expression
has no effect [-Werror=unused-value]
((hptr), (x)), 0)
Just removing the rh operand is no good, it will error in later:
linux-user/main.c: In function ‘arm_kernel_cmpxchg64_helper’:
linux-user/qemu.h:330:15: error: void value not ignored as it ought to be
__ret = __put_user((x), __hptr); \
Thus, remove setting __ret from __get_user and __put_user, as and
set the right hand operand to (void)0 to make it clear that these
return never nothing.
This commit depends on the signal.c cleanup, to ensure bisectable
version history.
Signed-off-by: Riku Voipio <riku.voipio@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Cc: Richard Henderson <rth@twiddle.net>
The last remaining check for return value of __get_user.
Signed-off-by: Riku Voipio <riku.voipio@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Cc: Alexander Graf <agraf@suse.de>
Remove checks of __get_user and the err variable
used to control flow with it.
Signed-off-by: Riku Voipio <riku.voipio@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
As __get_user and __put_user do not return errors, remove the
if checks from around them. This allows making the save/restore
functions void.
Signed-off-by: Riku Voipio <riku.voipio@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Cc: Alexander Graf <agraf@suse.de>
Remove "if(__put_user" checks and their related error paths
for all architecture's setup_frame, setup_rt_frame and similar.
Remove the unlock_user_struct when the only way to end up there is
from failed lock_user_struct.
Remove err variable if there are no users for it in the function
anymore.
Signed-off-by: Riku Voipio <riku.voipio@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Remove "if(__get_user" checks and their related error paths
for all architecture's do_sigreturn. Remove the unlock_user_struct
when the only way to end up there is from failed lock_user_struct.
v3: remove unneccesary sigsegv label as suggested by Peter
Signed-off-by: Riku Voipio <riku.voipio@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Access is already checked in the lock_user_struct
call before.
Signed-off-by: Riku Voipio <riku.voipio@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
A function never called from anywhere, obviously half-complete.
Remove function and if someone wants to complete this, please
check the old version out of git history.
Signed-off-by: Riku Voipio <riku.voipio@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
