On 04/12/2014 07:55 AM, Benjamin Herrenschmidt wrote:
> On Fri, 2014-04-11 at 11:40 +0200, Alexander Graf wrote:
>>> And G5 uses . I really do not understand why it is bad to
>>> send-and-check timer frequency. Why?
>>
>> Because the guest will continue to run at a different TB frequency on
Quoting Paolo Bonzini (pbonz...@redhat.com):
> Il 11/04/2014 04:31, Michael Tokarev ha scritto:
> >ENOENT means the kernel has an empty dirty bitmap for this
> >slot. Don't abort in that case. This appears to solve
> >the bug reported at
> >
> >https://bugs.launchpad.net/ubuntu/+source/qemu/+bug/
Il 11/04/2014 04:31, Michael Tokarev ha scritto:
ENOENT means the kernel has an empty dirty bitmap for this
slot. Don't abort in that case. This appears to solve
the bug reported at
https://bugs.launchpad.net/ubuntu/+source/qemu/+bug/1303926
which first showed up with commit b533f658a98325d:
Il 11/04/2014 13:37, Peter Maydell ha scritto:
Raised as issues but no patches:
* PCI bus naming
* win64 virtio-scsi regression
Stefano, can you post the patch to add the xen_enabled() check in
address_space_translate()?
Paolo
Il 07/04/2014 11:19, Michael S. Tsirkin ha scritto:
This means we can't cleanly implement an option for guest to
disable ACPI and switch to native if supported,
like the ACPI spec allows.
If we change hotplug code to walk the tree top down
and invoke all callbacks, then it will be fixed
in a cle
I was looking at rc2 and /scripts/qemu-binfmt-conf.sh briefly,
it has
# probe cpu type
cpu=`uname -m`
case "$cpu" in
Then there is no cpu for aarch64, but if you poke down further,
there is a if-then that uses a cpu=aarch64.
Should this be split into something like:
armv[4-7]*)
cpu="a
> Am 12.04.2014 um 00:59 schrieb Alexander Graf :
>
>
>
>> Am 11.04.2014 um 23:55 schrieb Benjamin Herrenschmidt
>> :
>>
>> On Fri, 2014-04-11 at 11:40 +0200, Alexander Graf wrote:
And G5 uses . I really do not understand why it is bad to
send-and-check timer frequency. Why
> Am 11.04.2014 um 23:55 schrieb Benjamin Herrenschmidt
> :
>
> On Fri, 2014-04-11 at 11:40 +0200, Alexander Graf wrote:
>>> And G5 uses . I really do not understand why it is bad to
>>> send-and-check timer frequency. Why?
>>
>> Because the guest will continue to run at a different TB
> Am 11.04.2014 um 19:37 schrieb Peter Maydell :
>
>> On 10 April 2014 12:17, Peter Maydell wrote:
>> So far I know of at least three fixes which should probably
>> go into 2.0
>
> Status update:
> Applied:
> * ACPI fixes (both sets)
> * block queue
> * SDL2 relative mode fixes
> * fix for vir
> Am 11.04.2014 um 19:05 schrieb Richard Henderson :
>
> On 04/11/2014 09:26 AM, Alexander Graf wrote:
>>>
>>> V2: updated patch 05/06 per Richard Henderson's review (const pointer to
>>> const)
>>
>> Looks good to me. Which queue should this go through? IMHO Richard would be
>> the
>> perfe
On Fri, Apr 11, 2014 at 11:41 PM, Markus Armbruster wrote:
> Peter Crosthwaite writes:
>
>> On Thu, Apr 10, 2014 at 1:48 AM, Markus Armbruster wrote:
>>> I stumbled over this while trying to purge error_is_set() from the code.
>>>
>>>
>>> Here's how we commonly use the Error API:
>>>
>>> Err
Signed-off-by: Cole Robinson
---
tests/.gitignore | 1 +
1 file changed, 1 insertion(+)
diff --git a/tests/.gitignore b/tests/.gitignore
index 3aa1e95..c71c110 100644
--- a/tests/.gitignore
+++ b/tests/.gitignore
@@ -23,6 +23,7 @@ test-qmp-input-strict
test-qmp-input-visitor
test-qmp-marshal.c
From: Laszlo Ersek
Also sort the test-* entries in the latter.
Signed-off-by: Laszlo Ersek
---
From: https://lists.gnu.org/archive/html/qemu-devel/2013-10/msg01998.html
v2:
Rebased to master
.gitignore | 9 -
tests/.gitignore | 11 +--
2 files changed, 9 insertions
But don't test their properties, otherwise we will recurse forever.
Their properties are already tested when we encounter them as child
properties elsewhere in the hierarchy, like /machine/unattached/...
This would have caught the crash fixed by 92b3eead
Signed-off-by: Cole Robinson
---
tests/q
On Fri, 2014-04-11 at 11:40 +0200, Alexander Graf wrote:
> > And G5 uses . I really do not understand why it is bad to
> > send-and-check timer frequency. Why?
>
> Because the guest will continue to run at a different TB frequency on
> the new host and break.
Right, which is why we shoul
Public bug reported:
Hi, I would like to report a possible error in the code
qemu/hw/net/opencores_eth.c
The corresponding data sheet :
http://www.cprover.org/firmware/doc/ethoc/eth_speci.pdf
In the code, there is a function open_eth_moder_host_write.
static void open_eth_moder_host_write(Ope
On 04/11/2014 09:28 AM, Takashi Iwai wrote:
> At Fri, 11 Apr 2014 15:07:34 +0200,
> Gerd Hoffmann wrote:
>
>> On Di, 2014-04-08 at 14:46 +0200, Takashi Iwai wrote:
>>> The current code shows annoying behavior where the X pointer can move
>>> out of the window during the input grab in the absolute
Works for me. I'll send a new series next week with this and Eric's corrections.
Lluis
Markus Armbruster writes:
> Lluís Vilanova writes:
>> Signed-off-by: Lluís Vilanova
>> ---
>> tests/qapi-schema/test-qapi.py |4 ++--
>> 1 file changed, 2 insertions(+), 2 deletions(-)
>>
>> diff --git
On 04/02/2014 09:16 AM, Lluís Vilanova wrote:
> Signed-off-by: Lluís Vilanova
> ---
It would be worth mentioning the new syntax in the commit message
proper. True, reading qapi-code-gen.txt covers the details, but
repeating (or summarizing) it in the commit log means I don't have to
leave 'git l
On Sun, 30 Mar 2014 11:58:22 +0100
Hani Benhabiles wrote:
> This patch series adds a new callback to mon_cmd_t which will make adding
> completion support for more commands cleaner.
>
> It then adds full or partial arguments completion for multiple hmp commands.
I was half-way through this seri
On Sun, 30 Mar 2014 11:58:30 +0100
Hani Benhabiles wrote:
> Also update the command's documentation.
>
> Signed-off-by: Hani Benhabiles
> ---
> hmp-commands.hx | 3 ++-
> hmp.h | 1 +
> monitor.c | 18 ++
> 3 files changed, 21 insertions(+), 1 deletion(-)
>
>
On Sun, 30 Mar 2014 11:58:29 +0100
Hani Benhabiles wrote:
> Make it possible to query all net clients without specifying an ID when
> calling
> qemu_find_net_clients_except().
>
> This also adds the add_completion_option() function which is to be used for
> other commands completions as well.
>
At the moment generic version-less CPUs are supported via hardcoded aliases.
For example, POWER7 is an alias for POWER7_v2.1. So when QEMU is started
with -cpu POWER7, the POWER7_v2.1 class instance is created.
This approach works for TCG and KVMs other than HV KVM. HV KVM cannot emulate
PVR value
On 10 April 2014 12:17, Peter Maydell wrote:
> So far I know of at least three fixes which should probably
> go into 2.0
Status update:
Applied:
* ACPI fixes (both sets)
* block queue
* SDL2 relative mode fixes
* fix for virtio-net CVE
* fix for qom-list crash
* my patch to stack-protector
On Sun, 30 Mar 2014 11:58:26 +0100
Hani Benhabiles wrote:
> Signed-off-by: Hani Benhabiles
Gerd, can you review this one and the patch (04/18). Both look good
to me, btw.
> ---
> hmp-commands.hx | 1 +
> hmp.h | 1 +
> monitor.c | 23 +++
> 3 files change
At the moment there are 3 versions of POWER7 CPUs defined. However
we do not emulate these CPUs diffent and it does not make much
sense to keep them all.
This removes POWER7_v2.0 and POWER7_v2.1 and leaves just one versioned
CPU per family which is POWER7_v2.3 with POWER7 alias.
Signed-off-by: Al
This enables generic CPU family name use with "-cpu" for all CPUs
across the family and not for the ones already defined in QEMU.
Changes:
v2:
* split one patch into 3
* added assert into ppc_cpu_get_family_class()
Alexey Kardashevskiy (3):
target-ppc: Create versionless CPU class per family i
This moves aliases lookup after CPU class lookup. This is to let new generic
CPU to be found first if it is present and only if it is not (TCG case), use
aliases.
Signed-off-by: Alexey Kardashevskiy
---
target-ppc/translate_init.c | 18 +++---
1 file changed, 11 insertions(+), 7 dele
On Sun, 30 Mar 2014 11:58:25 +0100
Hani Benhabiles wrote:
> Signed-off-by: Hani Benhabiles
This is not related to this change, but device_add seems to return
non-hotpluggable devices in the completion list (eg. Opteron_G5-x86_64-cpu).
Might be worth fixing in this series.
One more comment belo
On 04/11/2014 07:00 PM, Alexander Graf wrote:
>
> On 11.04.14 07:00, Alexey Kardashevskiy wrote:
>> At the moment generic version-less CPUs are supported via hardcoded aliases.
>> For example, POWER7 is an alias for POWER7_v2.1. So when QEMU is started
>> with -cpu POWER7, the POWER7_v2.1 class in
On Sun, 30 Mar 2014 11:58:24 +0100
Hani Benhabiles wrote:
> Convert object_add and object_del commands to use the new callback.
>
> Signed-off-by: Hani Benhabiles
> ---
> hmp-commands.hx | 2 ++
> hmp.h | 3 +++
> monitor.c | 19 +--
> 3 files changed, 18 inse
Rearrange code to put the compare and branch in the same place.
Reviewed-by: Claudio Fontana
Signed-off-by: Richard Henderson
---
tcg/aarch64/tcg-target.c | 34 ++
1 file changed, 14 insertions(+), 20 deletions(-)
diff --git a/tcg/aarch64/tcg-target.c b/tcg/aarc
This patch removes smbios_add_field() and the old code to insert
individual fields for types 0 and 1 into fw_cfg.
Signed-off-by: Gabriel Somlo
---
hw/i386/smbios.c | 80
1 file changed, 80 deletions(-)
diff --git a/hw/i386/smbios.c b/hw/i
Am 11.04.2014 19:05, schrieb Cole Robinson:
> On 04/11/2014 12:57 PM, Andreas Färber wrote:
>> Am 10.04.2014 20:47, schrieb Cole Robinson:
>>> Commit 9561fda8d90e176bef598ba87c42a1bd6ad03ef7 changed the type of
>>> 'opaque' for link properties, but missed updating this call site.
>>> Reproducer:
>>
Replace existing smbios_check_collision() functionality with
a pair of bitmaps: have_binfile_bitmap and have_fields_bitmap.
Bits corresponding to each smbios type are set by smbios_entry_add(),
which also uses the bitmaps to ensure that binary blobs and field
values are never accepted for the same
On 11 April 2014 17:57, Andreas Färber wrote:
> Am 10.04.2014 20:47, schrieb Cole Robinson:
>> Commit 9561fda8d90e176bef598ba87c42a1bd6ad03ef7 changed the type of
>> 'opaque' for link properties, but missed updating this call site.
>> Reproducer:
>>
>> ./x86_64-softmmu/qemu-system-x86_64 -qmp unix
This patch adds a set of macros which build full smbios tables
of a given type, including the logic to decide whether a given
table type should be built or not.
To illustrate this new functionality, we introduce and optionally
build a table of type 2 (base board), which is required by some
version
On 04/11/2014 09:26 AM, Alexander Graf wrote:
>>
>> V2: updated patch 05/06 per Richard Henderson's review (const pointer to
>> const)
>
> Looks good to me. Which queue should this go through? IMHO Richard would be
> the
> perfect person to get this through one of his.
I'm happy with it going t
On 04/11/2014 12:57 PM, Andreas Färber wrote:
> Am 10.04.2014 20:47, schrieb Cole Robinson:
>> Commit 9561fda8d90e176bef598ba87c42a1bd6ad03ef7 changed the type of
>> 'opaque' for link properties, but missed updating this call site.
>> Reproducer:
>>
>> ./x86_64-softmmu/qemu-system-x86_64 -qmp unix:
Build full tables for types 0 (bios information) and 1 (system
information). Type 0 is optional, and a table will only be built
if requested via the command line; the default is to leave type 0
tables up to the bios itself.
Signed-off-by: Gabriel Somlo
---
hw/i386/smbios.c | 63 +
Build type 19 (memory array mapped address, a.k.a. memory area)
tables by scanning the e820 map for E820_RAM entries. Since this
supercedes below_4g and above_4g ram amounts, we no longer need
them as arguments to smbios_set_defaults().
Signed-off-by: Gabriel Somlo
---
hw/i386/pc.c |
Am 10.04.2014 20:47, schrieb Cole Robinson:
> Commit 9561fda8d90e176bef598ba87c42a1bd6ad03ef7 changed the type of
> 'opaque' for link properties, but missed updating this call site.
> Reproducer:
>
> ./x86_64-softmmu/qemu-system-x86_64 -qmp unix:./qmp.sock,server &
> ./scripts/qmp/qmp-shell ./qmp.
Table definitions for types 4 and 17 are only up to v2.0,
so add fields specified in smbios v2.3, as expected (and
advertised) by the SeaBIOS smbios entry point structure.
In particular, OS X guests insist on type 17 being v2.3
compliant, to avoid GUI crashes when "about this mac" is
chosen in the
Some guest env are small enough to reach the tlb with only a 12-bit addition.
Reviewed-by: Claudio Fontana
Signed-off-by: Richard Henderson
---
tcg/aarch64/tcg-target.c | 28 +++-
1 file changed, 19 insertions(+), 9 deletions(-)
diff --git a/tcg/aarch64/tcg-target.c b/t
V4? The last patch set had a bizzare mix of v2/v3 markers...
Anyway, changes since last:
* Patch 1: Fix error in sigsegv write detection,
* Patch 21: Fix error in REV64; patch largely rewritten,
* Patch 23: Squashed 3 patches cleaning up loads; having these separate
didn't seem to help
- Replace some arbitrarily hardcoded fields with proper
"n/a" or "unknown" values;
- Use QEMU-supplied default manufacturer and version strings;
- Count CPUs starting with 0 instead of 1, to maintain uniformity
with other multiple-instance items.
Signed-off-by: Gabriel Somlo
---
hw
Check fw_cfg for the presence of a complete set of smbios tables
(etc/smbios/smbios-tables), and an entry point structure
(etc/smbios/smbios-anchor). If found, we no longer build smbios
tables locally; instead, we replace only the type 0 table with a
default of our own, and (re)calculate only the m
On 04/12/2014 02:29 AM, Alexander Graf wrote:
>
> On 13.03.14 07:29, Alexey Kardashevskiy wrote:
>> This adds a "ibm,chip-id" property for CPU nodes which should be the same
>> for all cores in the same CPU socket. The recent guest kernels use this
>> information to associate threads with sockets.
Instead of passing them the "args" array.
Reviewed-by: Claudio Fontana
Signed-off-by: Richard Henderson
---
tcg/aarch64/tcg-target.c | 49 +---
1 file changed, 17 insertions(+), 32 deletions(-)
diff --git a/tcg/aarch64/tcg-target.c b/tcg/aarch64/tcg-
On 10 April 2014 19:47, Cole Robinson wrote:
> Commit 9561fda8d90e176bef598ba87c42a1bd6ad03ef7 changed the type of
> 'opaque' for link properties, but missed updating this call site.
> Reproducer:
>
> ./x86_64-softmmu/qemu-system-x86_64 -qmp unix:./qmp.sock,server &
> ./scripts/qmp/qmp-shell ./qmp
On 04/12/2014 02:15 AM, Alexander Graf wrote:
>
> On 11.04.14 18:01, Alexey Kardashevskiy wrote:
>> On 04/12/2014 01:38 AM, Alexander Graf wrote:
>>> On 11.04.14 17:27, Alexey Kardashevskiy wrote:
On 04/12/2014 12:58 AM, Alexander Graf wrote:
> On 11.04.14 16:50, Alexey Kardashevskiy wrot
On 13.03.14 07:29, Alexey Kardashevskiy wrote:
This adds a "ibm,chip-id" property for CPU nodes which should be the same
for all cores in the same CPU socket. The recent guest kernels use this
information to associate threads with sockets.
Refer to the kernel commit 256f2d4b463d3030ebc8d2b54f42
Subsequent patches will utilize this function to set defaults for
more smbios types than just type 1, so the function name should
reflect this.
Signed-off-by: Gabriel Somlo
---
hw/i386/pc_piix.c| 12 ++--
hw/i386/pc_q35.c | 8
hw/i386/smbios.c | 4 ++--
On 04/12/2014 02:13 AM, Peter Maydell wrote:
> Since we use the -fstack-protector argument at both compile and
> link time in the build, we must check that it works with both
> a compile and a link:
> * MacOSX only fails in the compile step, not linking
> * some gcc cross environments only fail a
On 13.03.14 15:13, Tom Musta wrote:
This patch series addresses concerns raised by Richard Henderson regarding
redundant
copies of Advanced Encryption Standard (AES) data and code (see
http://lists.nongnu.org/archive/html/qemu-devel/2014-02/msg04391.html).
The patches declare commonly used AES
The function smbios_set_defaults() uses a repeating code pattern
for each field. This patch replaces that pattern with a macro.
Signed-off-by: Gabriel Somlo
---
hw/i386/smbios.c | 17 -
1 file changed, 8 insertions(+), 9 deletions(-)
diff --git a/hw/i386/smbios.c b/hw/i386/smbio
The subset of logical immediates that we support is quite quick to test,
and such constants are quite common to want to load.
Reviewed-by: Claudio Fontana
Signed-off-by: Richard Henderson
---
tcg/aarch64/tcg-target.c | 70 +++-
1 file changed, 39 inse
On 4/11/2014 10:31 AM, Alexander Graf wrote:
> I don't think this is possible. Libdecnumber is GPLv3+ code while QEMU
> overall is licensed at GPLv2 only. Gotta love licenses...
>
> Is there an older, GPLv2 licensed version available?
Interesting timing I was just about to put out a complet
Build a full set of smbios tables as a monolithic blob; Also,
build an entry point structure, and insert both the set
of tables and the entry point into distinct fw_cfg files.
This patch expects a SeaBIOS version equal or later than
commit X. An earlier
version will work, but will not be a
On 11.04.14 18:01, Alexey Kardashevskiy wrote:
On 04/12/2014 01:38 AM, Alexander Graf wrote:
On 11.04.14 17:27, Alexey Kardashevskiy wrote:
On 04/12/2014 12:58 AM, Alexander Graf wrote:
On 11.04.14 16:50, Alexey Kardashevskiy wrote:
On 04/11/2014 11:58 PM, Alexander Graf wrote:
On 11.04.201
Build full smbios type 32 (system boot info) and 127 (end-of-table)
tables, and make them available via fw_cfg.
Signed-off-by: Gabriel Somlo
---
hw/i386/smbios.c | 19 +++
1 file changed, 19 insertions(+)
diff --git a/hw/i386/smbios.c b/hw/i386/smbios.c
index 6510ff3..b1f1d46 10
On 11.04.14 18:12, Tom Musta wrote:
On 4/11/2014 10:31 AM, Alexander Graf wrote:
I don't think this is possible. Libdecnumber is GPLv3+ code while QEMU overall
is licensed at GPLv2 only. Gotta love licenses...
Is there an older, GPLv2 licensed version available?
Interesting timing I was
Table type 20 (memory device mapped address) is no longer required
as of smbios spec v2.5. Leaving it out completely saves us from having
to figure out how to connect type 17 devices to type 19 memory areas.
Signed-off-by: Gabriel Somlo
---
hw/i386/smbios.c | 99 +
Build full smbios tables representing the system RAM:
- type 16 (physical memory array): represents the entire system RAM;
- type 17 (memory device) tables: one per virtual DIMM;
- type 19 (memory array mapped address): represent major RAM areas
(currently one for below-4G memory, and, if
Since we use the -fstack-protector argument at both compile and
link time in the build, we must check that it works with both
a compile and a link:
* MacOSX only fails in the compile step, not linking
* some gcc cross environments only fail at the link stage (if they
require a libssp and it's
It's obviously call-clobbered, but is otherwise unused.
Repurpose it as the TCG temporary.
Reviewed-by: Claudio Fontana
Signed-off-by: Richard Henderson
---
tcg/aarch64/tcg-target.c | 34 --
tcg/aarch64/tcg-target.h | 32 +---
2 files
Build smbios type 3 (system enclosure) table, and make it available
to the bios via fw_cfg. For initial compatibility with SeaBIOS, use
"Bochs" as the default manufacturer string, and leave version unset.
Signed-off-by: Gabriel Somlo
---
hw/i386/smbios.c | 67
Build full smbios type 4 (processor information) tables, and make
them available to the bios via fw_cfg. For initial compatibility
with SeaBIOS, use "Bochs" as the default manufacturer string, and
leave version unset.
Signed-off-by: Gabriel Somlo
---
hw/i386/pc.c | 3 ++
hw/i386/smb
This patch adds extended start/end address and extended size fields
to each memory table type.
Signed-off-by: Gabriel Somlo
---
hw/i386/smbios.c | 46 +-
include/hw/i386/smbios.h | 15 ---
2 files changed, 41 insertions(+), 20 delet
With this update, we generate one type 4 (processor information) table
per socket, calculated as "smp_cpus / (smp_cores * smp_threads)", which
is in line with what happens on modern hardware.
Signed-off-by: Gabriel Somlo
---
hw/i386/smbios.c | 19 ++-
include/hw/i386/smbi
Signed-off-by: Gabriel Somlo
---
hw/i386/smbios.c | 8 +++-
include/hw/i386/smbios.h | 3 ++-
2 files changed, 9 insertions(+), 2 deletions(-)
diff --git a/hw/i386/smbios.c b/hw/i386/smbios.c
index b98d7ba..f5507cb 100644
--- a/hw/i386/smbios.c
+++ b/hw/i386/smbios.c
@@ -64,7 +64,7 @
A compare and branch against zero happens at the start of
every single TB.
Reviewed-by: Claudio Fontana
Signed-off-by: Richard Henderson
---
tcg/aarch64/tcg-target.c | 26 --
1 file changed, 24 insertions(+), 2 deletions(-)
diff --git a/tcg/aarch64/tcg-target.c b/tcg/aa
Rather than raw constants that could mean anything.
Reviewed-by: Claudio Fontana
Signed-off-by: Richard Henderson
---
tcg/aarch64/tcg-target.c | 73 +---
1 file changed, 38 insertions(+), 35 deletions(-)
diff --git a/tcg/aarch64/tcg-target.c b/tcg/aa
Quick highlights:
- after 10/18, we're generating tables 100% consistent with SeaBiOS
- after 14/18, we use e820 to generate type 19 (memory area) tables
- after 18/18, we expect SeaBIOS to check fw_cfg for a full set of
tables and an entry point (e.g., we need a version of SeaBIOS whi
On 04/12/2014 01:38 AM, Alexander Graf wrote:
>
> On 11.04.14 17:27, Alexey Kardashevskiy wrote:
>> On 04/12/2014 12:58 AM, Alexander Graf wrote:
>>> On 11.04.14 16:50, Alexey Kardashevskiy wrote:
On 04/11/2014 11:58 PM, Alexander Graf wrote:
> On 11.04.2014, at 14:38, Alexey Kardashevski
Quoting Michael Tokarev (m...@tls.msk.ru):
> ENOENT means the kernel has an empty dirty bitmap for this
> slot. Don't abort in that case. This appears to solve
> the bug reported at
>
> https://bugs.launchpad.net/ubuntu/+source/qemu/+bug/1303926
>
> which first showed up with commit b533f658a98
On 04/02/2014 09:16 AM, Lluís Vilanova wrote:
> Use an explicit input file on the command-line instead of reading from
> standard input
>
> Signed-off-by: Lluís Vilanova
> ---
> +input_file = ""
> output_dir = ""
> prefix = ""
> dispatch_type = "sync"
> @@ -389,6 +391,8 @@ do_h = False
> fo
Check fw_cfg for the presence of a complete set of smbios tables
(etc/smbios/smbios-tables), and an entry point structure
(etc/smbios/smbios-anchor). If found, no longer build smbios
tables locally; instead, replace only the type 0 table with a
default of our own, and (re)calculate only the minimum
It's the more canonical interface.
Reviewed-by: Claudio Fontana
Signed-off-by: Richard Henderson
---
tcg/aarch64/tcg-target.c | 16 +++-
1 file changed, 7 insertions(+), 9 deletions(-)
diff --git a/tcg/aarch64/tcg-target.c b/tcg/aarch64/tcg-target.c
index 7ff4be7..73ed658 100644
--
Replace aarch64_ldst_op_data with AArch64LdstType, as it wasn't encoded
for the proper shift for the field and was confusing.
Merge aarch64_ldst_op_data, AArch64LdstType, and a few stray opcode bits
into a single I3312_* argument, eliminating some magic numbers from the
helper functions.
Signed-o
Reviewed-by: Claudio Fontana
Signed-off-by: Richard Henderson
---
tcg/aarch64/tcg-target.c | 89
tcg/aarch64/tcg-target.h | 2 +-
2 files changed, 31 insertions(+), 60 deletions(-)
diff --git a/tcg/aarch64/tcg-target.c b/tcg/aarch64/tcg-target.c
Reviewed-by: Claudio Fontana
Signed-off-by: Richard Henderson
---
tcg/aarch64/tcg-target.c | 11 +--
1 file changed, 9 insertions(+), 2 deletions(-)
diff --git a/tcg/aarch64/tcg-target.c b/tcg/aarch64/tcg-target.c
index 4729d11..5d19e27 100644
--- a/tcg/aarch64/tcg-target.c
+++ b/tcg/aa
Reviewed-by: Claudio Fontana
Signed-off-by: Richard Henderson
---
tcg/aarch64/tcg-target.c | 35 +++
1 file changed, 19 insertions(+), 16 deletions(-)
diff --git a/tcg/aarch64/tcg-target.c b/tcg/aarch64/tcg-target.c
index 34e477d..caaf8a2 100644
--- a/tcg/aarch64
On 04/02/2014 09:16 AM, Lluís Vilanova wrote:
> Signed-off-by: Lluís Vilanova
> ---
> Makefile | 24 ++--
> tests/Makefile | 20
> 2 files changed, 34 insertions(+), 10 deletions(-)
Reviewed-by: Eric Blake
> + $(call quiet-command, PYTHONP
Making the bswap conditional on the memop instead of a compile-time test.
Reviewed-by: Claudio Fontana
Signed-off-by: Richard Henderson
---
tcg/aarch64/tcg-target.c | 131 +++
1 file changed, 63 insertions(+), 68 deletions(-)
diff --git a/tcg/aarch64
Reviewed-by: Claudio Fontana
Signed-off-by: Richard Henderson
---
tcg/aarch64/tcg-target.c | 84 +++-
1 file changed, 69 insertions(+), 15 deletions(-)
diff --git a/tcg/aarch64/tcg-target.c b/tcg/aarch64/tcg-target.c
index 5cffe50..4414bd1 100644
---
The assembler seems to prefer them, perhaps we should too.
Reviewed-by: Claudio Fontana
Signed-off-by: Richard Henderson
---
tcg/aarch64/tcg-target.c | 11 ++-
1 file changed, 6 insertions(+), 5 deletions(-)
diff --git a/tcg/aarch64/tcg-target.c b/tcg/aarch64/tcg-target.c
index 3235824
In some cases, a direct branch will be in range.
Reviewed-by: Claudio Fontana
Signed-off-by: Richard Henderson
---
tcg/aarch64/tcg-target.c | 6 ++
1 file changed, 2 insertions(+), 4 deletions(-)
diff --git a/tcg/aarch64/tcg-target.c b/tcg/aarch64/tcg-target.c
index 5186311..4729d11 100644
Reviewed-by: Claudio Fontana
Signed-off-by: Richard Henderson
---
tcg/aarch64/tcg-target.c | 74
1 file changed, 43 insertions(+), 31 deletions(-)
diff --git a/tcg/aarch64/tcg-target.c b/tcg/aarch64/tcg-target.c
index 1337a13..8b15d3b 100644
---
Reviewed-by: Claudio Fontana
Signed-off-by: Richard Henderson
---
tcg/aarch64/tcg-target.c | 115 +--
1 file changed, 32 insertions(+), 83 deletions(-)
diff --git a/tcg/aarch64/tcg-target.c b/tcg/aarch64/tcg-target.c
index 0846835..7f72df5 100644
---
Cleaning up the implementation of REV and REV16 at the same time.
Signed-off-by: Richard Henderson
---
tcg/aarch64/tcg-target.c | 57
1 file changed, 33 insertions(+), 24 deletions(-)
diff --git a/tcg/aarch64/tcg-target.c b/tcg/aarch64/tcg-target
Loading an qemu pointer as an immediate happens often. E.g.
- exit_tb $0x7fa8140013
+ exit_tb $0x7f81ee0013
...
- : d2800260mov x0, #0x13
- : f2b50280movkx0, #0xa814, lsl #16
- : f2c00fe0movkx0, #0x7f, lsl #32
+ : 90ff1000adrpx0, 0x7f81ee
+
Combines 4 other inline functions and tidies the prologue.
Reviewed-by: Claudio Fontana
Signed-off-by: Richard Henderson
---
tcg/aarch64/tcg-target.c | 100 ---
1 file changed, 33 insertions(+), 67 deletions(-)
diff --git a/tcg/aarch64/tcg-target.c b
Reviewed-by: Claudio Fontana
Signed-off-by: Richard Henderson
---
tcg/aarch64/tcg-target.c | 10 ++
1 file changed, 10 insertions(+)
diff --git a/tcg/aarch64/tcg-target.c b/tcg/aarch64/tcg-target.c
index c1d9895..a08f6c7 100644
--- a/tcg/aarch64/tcg-target.c
+++ b/tcg/aarch64/tcg-target
Since the kernel doesn't pass any info on the reason for the fault,
disassemble the instruction to detect a store.
Signed-off-by: Richard Henderson
---
user-exec.c | 29 +++--
1 file changed, 23 insertions(+), 6 deletions(-)
diff --git a/user-exec.c b/user-exec.c
index b
When profitable, initialize the register with MOVN instead of MOVZ,
before setting the remaining lanes with MOVK.
Reviewed-by: Claudio Fontana
Signed-off-by: Richard Henderson
---
tcg/aarch64/tcg-target.c | 63 ++--
1 file changed, 50 insertions(+), 1
As opposed to tcg_target_long.
Reviewed-by: Claudio Fontana
Signed-off-by: Richard Henderson
---
tcg/aarch64/tcg-target.c | 10 +-
1 file changed, 5 insertions(+), 5 deletions(-)
diff --git a/tcg/aarch64/tcg-target.c b/tcg/aarch64/tcg-target.c
index 661a5af..6938248 100644
--- a/tcg/aa
On 11 April 2014 15:21, Michael Tokarev wrote:
> 11.04.2014 16:18, Michael S. Tsirkin wrote:
>> When VM guest programs multicast addresses for
>> a virtio net card, it supplies a 32 bit
>> entries counter for the number of addresses.
>> These addresses are read into tail portion of
>> a fixed macs
On 11.04.14 17:27, Alexey Kardashevskiy wrote:
On 04/12/2014 12:58 AM, Alexander Graf wrote:
On 11.04.14 16:50, Alexey Kardashevskiy wrote:
On 04/11/2014 11:58 PM, Alexander Graf wrote:
On 11.04.2014, at 14:38, Alexey Kardashevskiy wrote:
On 04/11/2014 07:24 PM, Alexander Graf wrote:
On 1
Lluís Vilanova writes:
> Signed-off-by: Lluís Vilanova
> ---
> tests/qapi-schema/test-qapi.py |4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/tests/qapi-schema/test-qapi.py b/tests/qapi-schema/test-qapi.py
> index b3d1e1d..ac6da13 100644
> --- a/tests/qapi-schema
1 - 100 of 191 matches
Mail list logo