There's no need to attempt to match EXCP_* values with PALcode entry
point offsets. Instead, compress all the values to make for more
efficient switch statements within QEMU.
We will be doing TLB fill within QEMU proper, not within the PALcode,
so all of the ITB/DTB miss, double fault, and access exceptions can
be compressed to EXCP_MMFAULT.
Compress all of the EXCP_CALL_PAL exceptions into one.
Use env->error_code to store the specific entry point.
Signed-off-by: Richard Henderson <rth@twiddle.net>
With all of the pre-existing code that would not compile gone,
this is the earliest point at which the target can be enabled.
There is no machine defined yet, so this will crash on startup.
Enable the target anyway, to make sure that further compilation
problems do not creep back in.
Signed-off-by: Richard Henderson <rth@twiddle.net>
Delete all the code that tried to emulate the real IPRs of some
unnamed CPU. Replace those with just 3 slots that we can use to
communicate trap information between the helper functions that
signal exceptions and the OS trap handler.
Signed-off-by: Richard Henderson <rth@twiddle.net>
All of the "raw" memory accesses should be "phys" instead. Fix
some confusion about argument ordering of the store routines.
Fix the implementation of store-conditional.
Delete the "alt-mode" helpers. Because we only implement two
mmu modes, let /a imply user-mode unconditionally.
Leave some combinations of virt access without permission
checks as unimplemented. There are too many hoops through
which to jump, and these insns will not be needed in the
emulation palcode.
Signed-off-by: Richard Henderson <rth@twiddle.net>
We were failing to generate EXC_DEBUG in the EXIT_PC_UPDATED path.
This caused us not to stop at the instruction after a branch, but
on the instruction afterward.
Signed-off-by: Richard Henderson <rth@twiddle.net>
The QEMU emulation PALcode will use EV6 PALcode insns regardless
of the "real" cpu instruction set being emulated.
Signed-off-by: Richard Henderson <rth@twiddle.net>
ccid_initfn() allocates CCIDBus dynamically, but there is no exit
callback to free it.
Fix by getting rid of the allocation.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
No flag to configure is required. Instead, added a libcacard.la target that
is not built by default, only when requested explicitly via:
mkdir build
cd build
../configure
make libcacard.la
make install-libcacard
Uses libtool to do actual linking of object files and shared library, and
installing. Tested only under linux, but supposed to work on other systems as
well.
If libtool isn't found you get a message complaining about that, only at build
time (since it is not a default target I did not add a message at configure
time).
New build artifacts:
.libs subdirectories (at <buildroot> and <buildroot>/libcacard)
*.lo files (at same locations as the respective o files)
Added %.lo : %.c rule that uses libtool.
Updated clean rule to clean up those artifacts.
Added specific rule to call dtrace with libtool wrapper (note that because of
a current upstream dtrace bug fixed by systemtap b1568fd85 commit the -fPIC flag
isn't actually passed on. still current dtrace+libtool produced object links fine).
If libtool is missing any of the following targets will complain and exit 1:
any subdir: *.lo
root and libcacard: libcacard.la, libcacard-instsall
Tested to link and load with all tracing backends.
Thanks to Tobias Hoffmann <th55@gmx.de> for this patch.
Signed-off-by: Stefan Weil <weil@mail.berlios.de>
Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
After NACKing a read operation, a raising SCL should not trigger a new
read from the slave. Introduce a new state which just waits for a stop
or start condition after NACK.
Signed-off-by: Marcus Comstedt <marcus@mc.pp.se>
Signed-off-by: Andrzej Zaborowski <andrew.zaborowski@intel.com>
There is little in common with user and softmmu versions of cpu_resume_signal(),
split them.
Fix coding style for the user emulator part.
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
All you could ever achieve with it is break stuff, so removing it
should be safe.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>
virtio-serial-bus needs to treat "virtconsole" devices specially. It
uses VirtIOSerialPort member is_console to recognize them. It gets
its value via property initialization. Cute hack, except it lets
users mess with it: "-device virtconsole,is_console=0" isn't plugged
into port 0 as it should.
Move the flag to VirtIOSerialPortInfo. Keep the property for backward
compatibility; its value has no effect.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>
virtio_serial_init() allocates the VirtIOSerialBus dynamically, but
virtio_serial_exit() doesn't free it.
Fix by getting rid of the allocation.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>
Instead of calling flush_queued_data when unthrottling, schedule
a bh. That way we can return immediately to the caller, and the
flush uses the same call path as a have_data for callbackee.
No migration change is required because bh are called from vm_stop.
Signed-off-by: Alon Levy <alevy@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>
The LUN field in the CDB is a historical relic. Ignore it as reserved,
which is what modern SCSI specifications actually say.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
The sg driver currently has a hardcoded limit of commands it
can handle simultaneously. When this limit is reached the
driver will return -EDOM. So we need to capture this to
enable proper return values here.
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
scsi_req_parse() already provides for a data direction setting,
so we should be using it to check for correct direction.
And we should return the sense code 'INVALID FIELD IN CDB'
in these cases.
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
The get_sense callback copies existing sense information into
the provided buffer. This is required if sense information
should be transferred together with the command response.
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
... and remove some SCSIDevice variables or fields that now become unused.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Move the common part of scsi-disk.c and scsi-generic.c to the SCSI layer.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
The SCSI spec has a quite detailed list of sense codes available.
It even mandates the use of specific ones for some failure cases.
The current implementation just has one type of generic error
which is actually a violation of the spec in certain cases.
This patch introduces various predefined sense codes to have the
sense code reporting more in line with the spec.
On top of Hannes's patch I fixed the reply to REQUEST SENSE commands
with DESC=0 and a small (<18) length.
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
This is for when the request must be dropped in the void,
but still memory should be freed. To this end, the devices
register a second callback in SCSIBusOps.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
This covers the case of canceling a request's I/O and still
completing it.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
The code for canceling requests upon reset is already the same. Clean
it up and move it to scsi-bus.c.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Currently the SCSIRequest structure is abstracted away and cannot accessed
directly from the driver. This requires the handler to do a lookup on
an abstract 'tag' which identifies the SCSIRequest structure.
With this patch the SCSIRequest structure is exposed to the driver. This
allows use to use it directly as an argument to the SCSIDeviceInfo
callback functions and remove the lookup.
A new callback function 'alloc_req' is introduced matching 'free
req'; unref'ing to free up resources after use is moved into the
scsi_command_complete callbacks.
This temporarily introduces a leak of requests that are cancelled,
when they are removed from the queue and not from the driver. This
is fixed later by introducing scsi_req_cancel. That patch in turn
depends on this one, because the argument to scsi_req_cancel is a
SCSIRequest.
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
With the next patch, a device may hold SCSIRequest for an indefinite
time. Split a rather big patch, and protect against access errors,
by reference counting them.
There is some ugliness in scsi_send_command implementation due to
the need to unref the request when it fails. This will go away
with the next patches, which move the unref'ing to the devices.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
If a request is canceled after it has been completed, scsi_cancel_io
would pass a stale aiocb to bdrv_aio_cancel. Avoid this.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
