* [Qemu-devel] [PULL 01/30] configure: recognize more rpmbuild macros
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 02/30] cpus: Fix event order on resume of stopped guest Paolo Bonzini
` (29 subsequent siblings)
30 siblings, 0 replies; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel; +Cc: Olaf Hering, qemu-stable
From: Olaf Hering <olaf@aepfle.de>
Extend the list of recognized, but ignored options from rpms %configure
macro. This fixes build on hosts running SUSE Linux.
Cc: qemu-stable@nongnu.org
Signed-off-by: Olaf Hering <olaf@aepfle.de>
Message-Id: <20180418075045.27393-1-olaf@aepfle.de>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
configure | 2 ++
1 file changed, 2 insertions(+)
diff --git a/configure b/configure
index 1443422..83a6080 100755
--- a/configure
+++ b/configure
@@ -964,6 +964,8 @@ for opt do
;;
--firmwarepath=*) firmwarepath="$optarg"
;;
+ --host=*|--build=*|\
+ --disable-dependency-tracking|\
--sbindir=*|--sharedstatedir=*|\
--oldincludedir=*|--datarootdir=*|--infodir=*|--localedir=*|\
--htmldir=*|--dvidir=*|--pdfdir=*|--psdir=*)
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* [Qemu-devel] [PULL 02/30] cpus: Fix event order on resume of stopped guest
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 01/30] configure: recognize more rpmbuild macros Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 03/30] cpus: tcg: fix never exiting loop on unplug Paolo Bonzini
` (28 subsequent siblings)
30 siblings, 0 replies; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel; +Cc: Markus Armbruster
From: Markus Armbruster <armbru@redhat.com>
When resume of a stopped guest immediately runs into block device
errors, the BLOCK_IO_ERROR event is sent before the RESUME event.
Reproducer:
1. Create a scratch image
$ dd if=/dev/zero of=scratch.img bs=1M count=100
Size doesn't actually matter.
2. Prepare blkdebug configuration:
$ cat >blkdebug.conf <<EOF
[inject-error]
event = "write_aio"
errno = "5"
EOF
Note that errno 5 is EIO.
3. Run a guest with an additional scratch disk, i.e. with additional
arguments
-drive if=none,id=scratch-drive,format=raw,werror=stop,file=blkdebug:blkdebug.conf:scratch.img
-device virtio-blk-pci,id=scratch,drive=scratch-drive
The blkdebug part makes all writes to the scratch drive fail with
EIO. The werror=stop pauses the guest on write errors.
4. Connect to the QMP socket e.g. like this:
$ socat UNIX:/your/qmp/socket READLINE,history=$HOME/.qmp_history,prompt='QMP> '
Issue QMP command 'qmp_capabilities':
QMP> { "execute": "qmp_capabilities" }
5. Boot the guest.
6. In the guest, write to the scratch disk, e.g. like this:
# dd if=/dev/zero of=/dev/vdb count=1
Do double-check the device specified with of= is actually the
scratch device!
7. Issue QMP command 'cont':
QMP> { "execute": "cont" }
After step 6, I get a BLOCK_IO_ERROR event followed by a STOP event. Good.
After step 7, I get BLOCK_IO_ERROR, then RESUME, then STOP. Not so
good; I'd expect RESUME, then BLOCK_IO_ERROR, then STOP.
The funny event order confuses libvirt: virsh -r domstate DOMAIN
--reason reports "paused (unknown)" rather than "paused (I/O error)".
The culprit is vm_prepare_start().
/* Ensure that a STOP/RESUME pair of events is emitted if a
* vmstop request was pending. The BLOCK_IO_ERROR event, for
* example, according to documentation is always followed by
* the STOP event.
*/
if (runstate_is_running()) {
qapi_event_send_stop(&error_abort);
res = -1;
} else {
replay_enable_events();
cpu_enable_ticks();
runstate_set(RUN_STATE_RUNNING);
vm_state_notify(1, RUN_STATE_RUNNING);
}
/* We are sending this now, but the CPUs will be resumed shortly later */
qapi_event_send_resume(&error_abort);
return res;
When resuming a stopped guest, we take the else branch before we get
to sending RESUME. vm_state_notify() runs virtio_vmstate_change(),
among other things. This restarts I/O, triggering the BLOCK_IO_ERROR
event.
Reshuffle vm_prepare_start() to send the RESUME event earlier.
Fixes RHBZ 1566153.
Cc: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20180423084518.2426-1-armbru@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
cpus.c | 16 ++++++++--------
1 file changed, 8 insertions(+), 8 deletions(-)
diff --git a/cpus.c b/cpus.c
index 5bcd3ec..be3a4eb 100644
--- a/cpus.c
+++ b/cpus.c
@@ -2043,7 +2043,6 @@ int vm_stop(RunState state)
int vm_prepare_start(void)
{
RunState requested;
- int res = 0;
qemu_vmstop_requested(&requested);
if (runstate_is_running() && requested == RUN_STATE__MAX) {
@@ -2057,17 +2056,18 @@ int vm_prepare_start(void)
*/
if (runstate_is_running()) {
qapi_event_send_stop(&error_abort);
- res = -1;
- } else {
- replay_enable_events();
- cpu_enable_ticks();
- runstate_set(RUN_STATE_RUNNING);
- vm_state_notify(1, RUN_STATE_RUNNING);
+ qapi_event_send_resume(&error_abort);
+ return -1;
}
/* We are sending this now, but the CPUs will be resumed shortly later */
qapi_event_send_resume(&error_abort);
- return res;
+
+ replay_enable_events();
+ cpu_enable_ticks();
+ runstate_set(RUN_STATE_RUNNING);
+ vm_state_notify(1, RUN_STATE_RUNNING);
+ return 0;
}
void vm_start(void)
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* [Qemu-devel] [PULL 03/30] cpus: tcg: fix never exiting loop on unplug
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 01/30] configure: recognize more rpmbuild macros Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 02/30] cpus: Fix event order on resume of stopped guest Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 04/30] checkpatch.pl: add common glib defines to typelist Paolo Bonzini
` (27 subsequent siblings)
30 siblings, 0 replies; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel; +Cc: Cédric Le Goater, qemu-stable
From: Cédric Le Goater <clg@kaod.org>
Commit 9b0605f9837b ("cpus: tcg: unregister thread with RCU, fix
exiting of loop on unplug") changed the exit condition of the loop in
the vCPU thread function but forgot to remove the beginning 'while (1)'
statement. The resulting code :
while (1) {
...
} while (!cpu->unplug || cpu_can_run(cpu));
is a sequence of two distinct two while() loops, the first not exiting
in case of an unplug event.
Remove the first while (1) to fix CPU unplug.
Signed-off-by: Cédric Le Goater <clg@kaod.org>
Message-Id: <20180425131828.15604-1-clg@kaod.org>
Cc: qemu-stable@nongnu.org
Fixes: 9b0605f9837b68fd56c7fc7c96a3a1a3b983687d
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Cédric Le Goater <clg@kaod.org>
---
cpus.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/cpus.c b/cpus.c
index be3a4eb..d1f1629 100644
--- a/cpus.c
+++ b/cpus.c
@@ -1648,7 +1648,7 @@ static void *qemu_tcg_cpu_thread_fn(void *arg)
/* process any pending work */
cpu->exit_request = 1;
- while (1) {
+ do {
if (cpu_can_run(cpu)) {
int r;
qemu_mutex_unlock_iothread();
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* [Qemu-devel] [PULL 04/30] checkpatch.pl: add common glib defines to typelist
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
` (2 preceding siblings ...)
2018-05-08 22:14 ` [Qemu-devel] [PULL 03/30] cpus: tcg: fix never exiting loop on unplug Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 05/30] qom: allow object_get_canonical_path_component without parent Paolo Bonzini
` (26 subsequent siblings)
30 siblings, 0 replies; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel
Cc: Peter Xu, Daniel P. Berrangé,
Vladimir Sementsov-Ogievskiy, Fam Zheng
From: Peter Xu <peterx@redhat.com>
Otherwise it can warn this:
ERROR: space prohibited between function name and open parenthesis '('
When with things like this:
typedef gboolean (*it_tree_iterator)(ITValue start, ITValue end);
CC: Paolo Bonzini <pbonzini@redhat.com>
CC: "Daniel P. Berrangé" <berrange@redhat.com>
CC: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
CC: Fam Zheng <famz@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20180425070103.23723-1-peterx@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
scripts/checkpatch.pl | 28 ++++++++++++++++++++++++++++
1 file changed, 28 insertions(+)
diff --git a/scripts/checkpatch.pl b/scripts/checkpatch.pl
index 5b8735d..e73b4ef 100755
--- a/scripts/checkpatch.pl
+++ b/scripts/checkpatch.pl
@@ -265,8 +265,36 @@ our @typeList = (
qr{${Ident}_handler_fn},
qr{target_(?:u)?long},
qr{hwaddr},
+ # external libraries
qr{xml${Ident}},
qr{xendevicemodel_handle},
+ # Glib definitions
+ qr{gchar},
+ qr{gshort},
+ qr{glong},
+ qr{gint},
+ qr{gboolean},
+ qr{guchar},
+ qr{gushort},
+ qr{gulong},
+ qr{guint},
+ qr{gfloat},
+ qr{gdouble},
+ qr{gpointer},
+ qr{gconstpointer},
+ qr{gint8},
+ qr{guint8},
+ qr{gint16},
+ qr{guint16},
+ qr{gint32},
+ qr{guint32},
+ qr{gint64},
+ qr{guint64},
+ qr{gsize},
+ qr{gssize},
+ qr{goffset},
+ qr{gintptr},
+ qr{guintptr},
);
# This can be modified by sub possible. Since it can be empty, be careful
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* [Qemu-devel] [PULL 05/30] qom: allow object_get_canonical_path_component without parent
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
` (3 preceding siblings ...)
2018-05-08 22:14 ` [Qemu-devel] [PULL 04/30] checkpatch.pl: add common glib defines to typelist Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 06/30] memdev: remove "id" property Paolo Bonzini
` (25 subsequent siblings)
30 siblings, 0 replies; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel
Just return NULL; any callers that cause a change in behavior
would have caused an assertion failure before, so this is safe.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
include/qom/object.h | 1 +
qom/object.c | 5 +++--
2 files changed, 4 insertions(+), 2 deletions(-)
diff --git a/include/qom/object.h b/include/qom/object.h
index 96ce81b..a0c78c7 100644
--- a/include/qom/object.h
+++ b/include/qom/object.h
@@ -1302,6 +1302,7 @@ Object *object_get_internal_root(void);
*
* Returns: The final component in the object's canonical path. The canonical
* path is the path within the composition tree starting from the root.
+ * %NULL if the object doesn't have a parent (and thus a canonical path).
*/
gchar *object_get_canonical_path_component(Object *obj);
diff --git a/qom/object.c b/qom/object.c
index 76a89af..0fc9720 100644
--- a/qom/object.c
+++ b/qom/object.c
@@ -1644,8 +1644,9 @@ gchar *object_get_canonical_path_component(Object *obj)
ObjectProperty *prop = NULL;
GHashTableIter iter;
- g_assert(obj);
- g_assert(obj->parent != NULL);
+ if (obj->parent == NULL) {
+ return NULL;
+ }
g_hash_table_iter_init(&iter, obj->parent->properties);
while (g_hash_table_iter_next(&iter, NULL, (gpointer *)&prop)) {
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* [Qemu-devel] [PULL 06/30] memdev: remove "id" property
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
` (4 preceding siblings ...)
2018-05-08 22:14 ` [Qemu-devel] [PULL 05/30] qom: allow object_get_canonical_path_component without parent Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 07/30] exec: move memory access declarations to a common header, inline *_phys functions Paolo Bonzini
` (24 subsequent siblings)
30 siblings, 0 replies; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel
The "id" property is unnecessary and can be replaced simply with
object_get_canonical_path_component. This patch mostly undoes commit
e1ff3c67e8 ("monitor: fix qmp/hmp query-memdev not reporting IDs of
memory backends", 2017-01-12).
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
backends/hostmem.c | 26 --------------------------
include/sysemu/hostmem.h | 1 -
numa.c | 2 +-
qom/object_interfaces.c | 6 ------
4 files changed, 1 insertion(+), 34 deletions(-)
diff --git a/backends/hostmem.c b/backends/hostmem.c
index 6a0c474..3627e61 100644
--- a/backends/hostmem.c
+++ b/backends/hostmem.c
@@ -369,24 +369,6 @@ host_memory_backend_can_be_deleted(UserCreatable *uc)
}
}
-static char *get_id(Object *o, Error **errp)
-{
- HostMemoryBackend *backend = MEMORY_BACKEND(o);
-
- return g_strdup(backend->id);
-}
-
-static void set_id(Object *o, const char *str, Error **errp)
-{
- HostMemoryBackend *backend = MEMORY_BACKEND(o);
-
- if (backend->id) {
- error_setg(errp, "cannot change property value");
- return;
- }
- backend->id = g_strdup(str);
-}
-
static bool host_memory_backend_get_share(Object *o, Error **errp)
{
HostMemoryBackend *backend = MEMORY_BACKEND(o);
@@ -434,18 +416,11 @@ host_memory_backend_class_init(ObjectClass *oc, void *data)
&HostMemPolicy_lookup,
host_memory_backend_get_policy,
host_memory_backend_set_policy, &error_abort);
- object_class_property_add_str(oc, "id", get_id, set_id, &error_abort);
object_class_property_add_bool(oc, "share",
host_memory_backend_get_share, host_memory_backend_set_share,
&error_abort);
}
-static void host_memory_backend_finalize(Object *o)
-{
- HostMemoryBackend *backend = MEMORY_BACKEND(o);
- g_free(backend->id);
-}
-
static const TypeInfo host_memory_backend_info = {
.name = TYPE_MEMORY_BACKEND,
.parent = TYPE_OBJECT,
@@ -454,7 +429,6 @@ static const TypeInfo host_memory_backend_info = {
.class_init = host_memory_backend_class_init,
.instance_size = sizeof(HostMemoryBackend),
.instance_init = host_memory_backend_init,
- .instance_finalize = host_memory_backend_finalize,
.interfaces = (InterfaceInfo[]) {
{ TYPE_USER_CREATABLE },
{ }
diff --git a/include/sysemu/hostmem.h b/include/sysemu/hostmem.h
index bc36899..5beb0ef 100644
--- a/include/sysemu/hostmem.h
+++ b/include/sysemu/hostmem.h
@@ -52,7 +52,6 @@ struct HostMemoryBackend {
Object parent;
/* protected */
- char *id;
uint64_t size;
bool merge, dump;
bool prealloc, force_prealloc, is_mapped, share;
diff --git a/numa.c b/numa.c
index 70b150e..aac22a9 100644
--- a/numa.c
+++ b/numa.c
@@ -580,7 +580,7 @@ static int query_memdev(Object *obj, void *opaque)
m->value = g_malloc0(sizeof(*m->value));
- m->value->id = object_property_get_str(obj, "id", NULL);
+ m->value->id = object_get_canonical_path_component(obj);
m->value->has_id = !!m->value->id;
m->value->size = object_property_get_uint(obj, "size",
diff --git a/qom/object_interfaces.c b/qom/object_interfaces.c
index 980ffc2..72b97a8 100644
--- a/qom/object_interfaces.c
+++ b/qom/object_interfaces.c
@@ -65,12 +65,6 @@ Object *user_creatable_add_type(const char *type, const char *id,
assert(qdict);
obj = object_new(type);
- if (object_property_find(obj, "id", NULL)) {
- object_property_set_str(obj, id, "id", &local_err);
- if (local_err) {
- goto out;
- }
- }
visit_start_struct(v, NULL, NULL, 0, &local_err);
if (local_err) {
goto out;
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* [Qemu-devel] [PULL 07/30] exec: move memory access declarations to a common header, inline *_phys functions
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
` (5 preceding siblings ...)
2018-05-08 22:14 ` [Qemu-devel] [PULL 06/30] memdev: remove "id" property Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 08/30] exec: small changes to flatview_do_translate Paolo Bonzini
` (23 subsequent siblings)
30 siblings, 0 replies; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel
For now, this reduces the text size very slightly due to the newly-added
inlining:
text size before: 9301965
text size after: 9300645
Later, however, the declarations in include/exec/memory_ldst.inc.h will be
reused for the MemoryRegionCache slow path functions.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
include/exec/cpu-all.h | 75 +++++++-----------
include/exec/memory.h | 153 +++++++++++-------------------------
include/exec/memory_ldst.inc.h | 71 +++++++++++++++++
include/exec/memory_ldst_phys.inc.h | 147 ++++++++++++++++++++++++++++++++++
memory_ldst.inc.c | 126 -----------------------------
5 files changed, 292 insertions(+), 280 deletions(-)
create mode 100644 include/exec/memory_ldst.inc.h
create mode 100644 include/exec/memory_ldst_phys.inc.h
diff --git a/include/exec/cpu-all.h b/include/exec/cpu-all.h
index f4fa94e..173edd1 100644
--- a/include/exec/cpu-all.h
+++ b/include/exec/cpu-all.h
@@ -168,51 +168,36 @@ extern unsigned long reserved_va;
#else
#include "exec/hwaddr.h"
-uint32_t lduw_phys(AddressSpace *as, hwaddr addr);
-uint32_t ldl_phys(AddressSpace *as, hwaddr addr);
-uint64_t ldq_phys(AddressSpace *as, hwaddr addr);
-void stl_phys_notdirty(AddressSpace *as, hwaddr addr, uint32_t val);
-void stw_phys(AddressSpace *as, hwaddr addr, uint32_t val);
-void stl_phys(AddressSpace *as, hwaddr addr, uint32_t val);
-void stq_phys(AddressSpace *as, hwaddr addr, uint64_t val);
-
-uint32_t address_space_lduw(AddressSpace *as, hwaddr addr,
- MemTxAttrs attrs, MemTxResult *result);
-uint32_t address_space_ldl(AddressSpace *as, hwaddr addr,
- MemTxAttrs attrs, MemTxResult *result);
-uint64_t address_space_ldq(AddressSpace *as, hwaddr addr,
- MemTxAttrs attrs, MemTxResult *result);
-void address_space_stl_notdirty(AddressSpace *as, hwaddr addr, uint32_t val,
- MemTxAttrs attrs, MemTxResult *result);
-void address_space_stw(AddressSpace *as, hwaddr addr, uint32_t val,
- MemTxAttrs attrs, MemTxResult *result);
-void address_space_stl(AddressSpace *as, hwaddr addr, uint32_t val,
- MemTxAttrs attrs, MemTxResult *result);
-void address_space_stq(AddressSpace *as, hwaddr addr, uint64_t val,
- MemTxAttrs attrs, MemTxResult *result);
-
-uint32_t lduw_phys_cached(MemoryRegionCache *cache, hwaddr addr);
-uint32_t ldl_phys_cached(MemoryRegionCache *cache, hwaddr addr);
-uint64_t ldq_phys_cached(MemoryRegionCache *cache, hwaddr addr);
-void stl_phys_notdirty_cached(MemoryRegionCache *cache, hwaddr addr, uint32_t val);
-void stw_phys_cached(MemoryRegionCache *cache, hwaddr addr, uint32_t val);
-void stl_phys_cached(MemoryRegionCache *cache, hwaddr addr, uint32_t val);
-void stq_phys_cached(MemoryRegionCache *cache, hwaddr addr, uint64_t val);
-
-uint32_t address_space_lduw_cached(MemoryRegionCache *cache, hwaddr addr,
- MemTxAttrs attrs, MemTxResult *result);
-uint32_t address_space_ldl_cached(MemoryRegionCache *cache, hwaddr addr,
- MemTxAttrs attrs, MemTxResult *result);
-uint64_t address_space_ldq_cached(MemoryRegionCache *cache, hwaddr addr,
- MemTxAttrs attrs, MemTxResult *result);
-void address_space_stl_notdirty_cached(MemoryRegionCache *cache, hwaddr addr,
- uint32_t val, MemTxAttrs attrs, MemTxResult *result);
-void address_space_stw_cached(MemoryRegionCache *cache, hwaddr addr, uint32_t val,
- MemTxAttrs attrs, MemTxResult *result);
-void address_space_stl_cached(MemoryRegionCache *cache, hwaddr addr, uint32_t val,
- MemTxAttrs attrs, MemTxResult *result);
-void address_space_stq_cached(MemoryRegionCache *cache, hwaddr addr, uint64_t val,
- MemTxAttrs attrs, MemTxResult *result);
+
+#define SUFFIX
+#define ARG1 as
+#define ARG1_DECL AddressSpace *as
+#define TARGET_ENDIANNESS
+#include "exec/memory_ldst.inc.h"
+
+#define SUFFIX _cached
+#define ARG1 cache
+#define ARG1_DECL MemoryRegionCache *cache
+#define TARGET_ENDIANNESS
+#include "exec/memory_ldst.inc.h"
+
+static inline void stl_phys_notdirty(AddressSpace *as, hwaddr addr, uint32_t val)
+{
+ address_space_stl_notdirty(as, addr, val,
+ MEMTXATTRS_UNSPECIFIED, NULL);
+}
+
+#define SUFFIX
+#define ARG1 as
+#define ARG1_DECL AddressSpace *as
+#define TARGET_ENDIANNESS
+#include "exec/memory_ldst_phys.inc.h"
+
+#define SUFFIX _cached
+#define ARG1 cache
+#define ARG1_DECL MemoryRegionCache *cache
+#define TARGET_ENDIANNESS
+#include "exec/memory_ldst_phys.inc.h"
#endif
/* page related stuff */
diff --git a/include/exec/memory.h b/include/exec/memory.h
index 31eae0a..ca361bc 100644
--- a/include/exec/memory.h
+++ b/include/exec/memory.h
@@ -1676,49 +1676,16 @@ MemTxResult address_space_write(AddressSpace *as, hwaddr addr,
* @result: location to write the success/failure of the transaction;
* if NULL, this information is discarded
*/
-uint32_t address_space_ldub(AddressSpace *as, hwaddr addr,
- MemTxAttrs attrs, MemTxResult *result);
-uint32_t address_space_lduw_le(AddressSpace *as, hwaddr addr,
- MemTxAttrs attrs, MemTxResult *result);
-uint32_t address_space_lduw_be(AddressSpace *as, hwaddr addr,
- MemTxAttrs attrs, MemTxResult *result);
-uint32_t address_space_ldl_le(AddressSpace *as, hwaddr addr,
- MemTxAttrs attrs, MemTxResult *result);
-uint32_t address_space_ldl_be(AddressSpace *as, hwaddr addr,
- MemTxAttrs attrs, MemTxResult *result);
-uint64_t address_space_ldq_le(AddressSpace *as, hwaddr addr,
- MemTxAttrs attrs, MemTxResult *result);
-uint64_t address_space_ldq_be(AddressSpace *as, hwaddr addr,
- MemTxAttrs attrs, MemTxResult *result);
-void address_space_stb(AddressSpace *as, hwaddr addr, uint32_t val,
- MemTxAttrs attrs, MemTxResult *result);
-void address_space_stw_le(AddressSpace *as, hwaddr addr, uint32_t val,
- MemTxAttrs attrs, MemTxResult *result);
-void address_space_stw_be(AddressSpace *as, hwaddr addr, uint32_t val,
- MemTxAttrs attrs, MemTxResult *result);
-void address_space_stl_le(AddressSpace *as, hwaddr addr, uint32_t val,
- MemTxAttrs attrs, MemTxResult *result);
-void address_space_stl_be(AddressSpace *as, hwaddr addr, uint32_t val,
- MemTxAttrs attrs, MemTxResult *result);
-void address_space_stq_le(AddressSpace *as, hwaddr addr, uint64_t val,
- MemTxAttrs attrs, MemTxResult *result);
-void address_space_stq_be(AddressSpace *as, hwaddr addr, uint64_t val,
- MemTxAttrs attrs, MemTxResult *result);
-
-uint32_t ldub_phys(AddressSpace *as, hwaddr addr);
-uint32_t lduw_le_phys(AddressSpace *as, hwaddr addr);
-uint32_t lduw_be_phys(AddressSpace *as, hwaddr addr);
-uint32_t ldl_le_phys(AddressSpace *as, hwaddr addr);
-uint32_t ldl_be_phys(AddressSpace *as, hwaddr addr);
-uint64_t ldq_le_phys(AddressSpace *as, hwaddr addr);
-uint64_t ldq_be_phys(AddressSpace *as, hwaddr addr);
-void stb_phys(AddressSpace *as, hwaddr addr, uint32_t val);
-void stw_le_phys(AddressSpace *as, hwaddr addr, uint32_t val);
-void stw_be_phys(AddressSpace *as, hwaddr addr, uint32_t val);
-void stl_le_phys(AddressSpace *as, hwaddr addr, uint32_t val);
-void stl_be_phys(AddressSpace *as, hwaddr addr, uint32_t val);
-void stq_le_phys(AddressSpace *as, hwaddr addr, uint64_t val);
-void stq_be_phys(AddressSpace *as, hwaddr addr, uint64_t val);
+
+#define SUFFIX
+#define ARG1 as
+#define ARG1_DECL AddressSpace *as
+#include "exec/memory_ldst.inc.h"
+
+#define SUFFIX
+#define ARG1 as
+#define ARG1_DECL AddressSpace *as
+#include "exec/memory_ldst_phys.inc.h"
struct MemoryRegionCache {
hwaddr xlat;
@@ -1728,6 +1695,40 @@ struct MemoryRegionCache {
#define MEMORY_REGION_CACHE_INVALID ((MemoryRegionCache) { .as = NULL })
+/* address_space_ld*_cached: load from a cached #MemoryRegion
+ * address_space_st*_cached: store into a cached #MemoryRegion
+ *
+ * These functions perform a load or store of the byte, word,
+ * longword or quad to the specified address. The address is
+ * a physical address in the AddressSpace, but it must lie within
+ * a #MemoryRegion that was mapped with address_space_cache_init.
+ *
+ * The _le suffixed functions treat the data as little endian;
+ * _be indicates big endian; no suffix indicates "same endianness
+ * as guest CPU".
+ *
+ * The "guest CPU endianness" accessors are deprecated for use outside
+ * target-* code; devices should be CPU-agnostic and use either the LE
+ * or the BE accessors.
+ *
+ * @cache: previously initialized #MemoryRegionCache to be accessed
+ * @addr: address within the address space
+ * @val: data value, for stores
+ * @attrs: memory transaction attributes
+ * @result: location to write the success/failure of the transaction;
+ * if NULL, this information is discarded
+ */
+
+#define SUFFIX _cached
+#define ARG1 cache
+#define ARG1_DECL MemoryRegionCache *cache
+#include "exec/memory_ldst.inc.h"
+
+#define SUFFIX _cached
+#define ARG1 cache
+#define ARG1_DECL MemoryRegionCache *cache
+#include "exec/memory_ldst_phys.inc.h"
+
/* address_space_cache_init: prepare for repeated access to a physical
* memory region
*
@@ -1772,72 +1773,6 @@ void address_space_cache_invalidate(MemoryRegionCache *cache,
*/
void address_space_cache_destroy(MemoryRegionCache *cache);
-/* address_space_ld*_cached: load from a cached #MemoryRegion
- * address_space_st*_cached: store into a cached #MemoryRegion
- *
- * These functions perform a load or store of the byte, word,
- * longword or quad to the specified address. The address is
- * a physical address in the AddressSpace, but it must lie within
- * a #MemoryRegion that was mapped with address_space_cache_init.
- *
- * The _le suffixed functions treat the data as little endian;
- * _be indicates big endian; no suffix indicates "same endianness
- * as guest CPU".
- *
- * The "guest CPU endianness" accessors are deprecated for use outside
- * target-* code; devices should be CPU-agnostic and use either the LE
- * or the BE accessors.
- *
- * @cache: previously initialized #MemoryRegionCache to be accessed
- * @addr: address within the address space
- * @val: data value, for stores
- * @attrs: memory transaction attributes
- * @result: location to write the success/failure of the transaction;
- * if NULL, this information is discarded
- */
-uint32_t address_space_ldub_cached(MemoryRegionCache *cache, hwaddr addr,
- MemTxAttrs attrs, MemTxResult *result);
-uint32_t address_space_lduw_le_cached(MemoryRegionCache *cache, hwaddr addr,
- MemTxAttrs attrs, MemTxResult *result);
-uint32_t address_space_lduw_be_cached(MemoryRegionCache *cache, hwaddr addr,
- MemTxAttrs attrs, MemTxResult *result);
-uint32_t address_space_ldl_le_cached(MemoryRegionCache *cache, hwaddr addr,
- MemTxAttrs attrs, MemTxResult *result);
-uint32_t address_space_ldl_be_cached(MemoryRegionCache *cache, hwaddr addr,
- MemTxAttrs attrs, MemTxResult *result);
-uint64_t address_space_ldq_le_cached(MemoryRegionCache *cache, hwaddr addr,
- MemTxAttrs attrs, MemTxResult *result);
-uint64_t address_space_ldq_be_cached(MemoryRegionCache *cache, hwaddr addr,
- MemTxAttrs attrs, MemTxResult *result);
-void address_space_stb_cached(MemoryRegionCache *cache, hwaddr addr, uint32_t val,
- MemTxAttrs attrs, MemTxResult *result);
-void address_space_stw_le_cached(MemoryRegionCache *cache, hwaddr addr, uint32_t val,
- MemTxAttrs attrs, MemTxResult *result);
-void address_space_stw_be_cached(MemoryRegionCache *cache, hwaddr addr, uint32_t val,
- MemTxAttrs attrs, MemTxResult *result);
-void address_space_stl_le_cached(MemoryRegionCache *cache, hwaddr addr, uint32_t val,
- MemTxAttrs attrs, MemTxResult *result);
-void address_space_stl_be_cached(MemoryRegionCache *cache, hwaddr addr, uint32_t val,
- MemTxAttrs attrs, MemTxResult *result);
-void address_space_stq_le_cached(MemoryRegionCache *cache, hwaddr addr, uint64_t val,
- MemTxAttrs attrs, MemTxResult *result);
-void address_space_stq_be_cached(MemoryRegionCache *cache, hwaddr addr, uint64_t val,
- MemTxAttrs attrs, MemTxResult *result);
-
-uint32_t ldub_phys_cached(MemoryRegionCache *cache, hwaddr addr);
-uint32_t lduw_le_phys_cached(MemoryRegionCache *cache, hwaddr addr);
-uint32_t lduw_be_phys_cached(MemoryRegionCache *cache, hwaddr addr);
-uint32_t ldl_le_phys_cached(MemoryRegionCache *cache, hwaddr addr);
-uint32_t ldl_be_phys_cached(MemoryRegionCache *cache, hwaddr addr);
-uint64_t ldq_le_phys_cached(MemoryRegionCache *cache, hwaddr addr);
-uint64_t ldq_be_phys_cached(MemoryRegionCache *cache, hwaddr addr);
-void stb_phys_cached(MemoryRegionCache *cache, hwaddr addr, uint32_t val);
-void stw_le_phys_cached(MemoryRegionCache *cache, hwaddr addr, uint32_t val);
-void stw_be_phys_cached(MemoryRegionCache *cache, hwaddr addr, uint32_t val);
-void stl_le_phys_cached(MemoryRegionCache *cache, hwaddr addr, uint32_t val);
-void stl_be_phys_cached(MemoryRegionCache *cache, hwaddr addr, uint32_t val);
-void stq_le_phys_cached(MemoryRegionCache *cache, hwaddr addr, uint64_t val);
-void stq_be_phys_cached(MemoryRegionCache *cache, hwaddr addr, uint64_t val);
/* address_space_get_iotlb_entry: translate an address into an IOTLB
* entry. Should be called from an RCU critical section.
*/
diff --git a/include/exec/memory_ldst.inc.h b/include/exec/memory_ldst.inc.h
new file mode 100644
index 0000000..272c20f
--- /dev/null
+++ b/include/exec/memory_ldst.inc.h
@@ -0,0 +1,71 @@
+/*
+ * Physical memory access templates
+ *
+ * Copyright (c) 2003 Fabrice Bellard
+ * Copyright (c) 2015 Linaro, Inc.
+ * Copyright (c) 2016 Red Hat, Inc.
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library; if not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifdef TARGET_ENDIANNESS
+extern uint32_t glue(address_space_lduw, SUFFIX)(ARG1_DECL,
+ hwaddr addr, MemTxAttrs attrs, MemTxResult *result);
+extern uint32_t glue(address_space_ldl, SUFFIX)(ARG1_DECL,
+ hwaddr addr, MemTxAttrs attrs, MemTxResult *result);
+extern uint64_t glue(address_space_ldq, SUFFIX)(ARG1_DECL,
+ hwaddr addr, MemTxAttrs attrs, MemTxResult *result);
+extern void glue(address_space_stl_notdirty, SUFFIX)(ARG1_DECL,
+ hwaddr addr, uint32_t val, MemTxAttrs attrs, MemTxResult *result);
+extern void glue(address_space_stw, SUFFIX)(ARG1_DECL,
+ hwaddr addr, uint32_t val, MemTxAttrs attrs, MemTxResult *result);
+extern void glue(address_space_stl, SUFFIX)(ARG1_DECL,
+ hwaddr addr, uint32_t val, MemTxAttrs attrs, MemTxResult *result);
+extern void glue(address_space_stq, SUFFIX)(ARG1_DECL,
+ hwaddr addr, uint64_t val, MemTxAttrs attrs, MemTxResult *result);
+#else
+extern uint32_t glue(address_space_ldub, SUFFIX)(ARG1_DECL,
+ hwaddr addr, MemTxAttrs attrs, MemTxResult *result);
+extern uint32_t glue(address_space_lduw_le, SUFFIX)(ARG1_DECL,
+ hwaddr addr, MemTxAttrs attrs, MemTxResult *result);
+extern uint32_t glue(address_space_lduw_be, SUFFIX)(ARG1_DECL,
+ hwaddr addr, MemTxAttrs attrs, MemTxResult *result);
+extern uint32_t glue(address_space_ldl_le, SUFFIX)(ARG1_DECL,
+ hwaddr addr, MemTxAttrs attrs, MemTxResult *result);
+extern uint32_t glue(address_space_ldl_be, SUFFIX)(ARG1_DECL,
+ hwaddr addr, MemTxAttrs attrs, MemTxResult *result);
+extern uint64_t glue(address_space_ldq_le, SUFFIX)(ARG1_DECL,
+ hwaddr addr, MemTxAttrs attrs, MemTxResult *result);
+extern uint64_t glue(address_space_ldq_be, SUFFIX)(ARG1_DECL,
+ hwaddr addr, MemTxAttrs attrs, MemTxResult *result);
+extern void glue(address_space_stb, SUFFIX)(ARG1_DECL,
+ hwaddr addr, uint32_t val, MemTxAttrs attrs, MemTxResult *result);
+extern void glue(address_space_stw_le, SUFFIX)(ARG1_DECL,
+ hwaddr addr, uint32_t val, MemTxAttrs attrs, MemTxResult *result);
+extern void glue(address_space_stw_be, SUFFIX)(ARG1_DECL,
+ hwaddr addr, uint32_t val, MemTxAttrs attrs, MemTxResult *result);
+extern void glue(address_space_stl_le, SUFFIX)(ARG1_DECL,
+ hwaddr addr, uint32_t val, MemTxAttrs attrs, MemTxResult *result);
+extern void glue(address_space_stl_be, SUFFIX)(ARG1_DECL,
+ hwaddr addr, uint32_t val, MemTxAttrs attrs, MemTxResult *result);
+extern void glue(address_space_stq_le, SUFFIX)(ARG1_DECL,
+ hwaddr addr, uint64_t val, MemTxAttrs attrs, MemTxResult *result);
+extern void glue(address_space_stq_be, SUFFIX)(ARG1_DECL,
+ hwaddr addr, uint64_t val, MemTxAttrs attrs, MemTxResult *result);
+#endif
+
+#undef ARG1_DECL
+#undef ARG1
+#undef SUFFIX
+#undef TARGET_ENDIANNESS
diff --git a/include/exec/memory_ldst_phys.inc.h b/include/exec/memory_ldst_phys.inc.h
new file mode 100644
index 0000000..91f7297
--- /dev/null
+++ b/include/exec/memory_ldst_phys.inc.h
@@ -0,0 +1,147 @@
+/*
+ * Physical memory access templates
+ *
+ * Copyright (c) 2003 Fabrice Bellard
+ * Copyright (c) 2015 Linaro, Inc.
+ * Copyright (c) 2016 Red Hat, Inc.
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library; if not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifdef TARGET_ENDIANNESS
+static inline uint32_t glue(ldl_phys, SUFFIX)(ARG1_DECL, hwaddr addr)
+{
+ return glue(address_space_ldl, SUFFIX)(ARG1, addr,
+ MEMTXATTRS_UNSPECIFIED, NULL);
+}
+
+static inline uint64_t glue(ldq_phys, SUFFIX)(ARG1_DECL, hwaddr addr)
+{
+ return glue(address_space_ldq, SUFFIX)(ARG1, addr,
+ MEMTXATTRS_UNSPECIFIED, NULL);
+}
+
+static inline uint32_t glue(lduw_phys, SUFFIX)(ARG1_DECL, hwaddr addr)
+{
+ return glue(address_space_lduw, SUFFIX)(ARG1, addr,
+ MEMTXATTRS_UNSPECIFIED, NULL);
+}
+
+static inline void glue(stl_phys, SUFFIX)(ARG1_DECL, hwaddr addr, uint32_t val)
+{
+ glue(address_space_stl, SUFFIX)(ARG1, addr, val,
+ MEMTXATTRS_UNSPECIFIED, NULL);
+}
+
+static inline void glue(stw_phys, SUFFIX)(ARG1_DECL, hwaddr addr, uint32_t val)
+{
+ glue(address_space_stw, SUFFIX)(ARG1, addr, val,
+ MEMTXATTRS_UNSPECIFIED, NULL);
+}
+
+static inline void glue(stq_phys, SUFFIX)(ARG1_DECL, hwaddr addr, uint64_t val)
+{
+ glue(address_space_stq, SUFFIX)(ARG1, addr, val,
+ MEMTXATTRS_UNSPECIFIED, NULL);
+}
+#else
+static inline uint32_t glue(ldl_le_phys, SUFFIX)(ARG1_DECL, hwaddr addr)
+{
+ return glue(address_space_ldl_le, SUFFIX)(ARG1, addr,
+ MEMTXATTRS_UNSPECIFIED, NULL);
+}
+
+static inline uint32_t glue(ldl_be_phys, SUFFIX)(ARG1_DECL, hwaddr addr)
+{
+ return glue(address_space_ldl_be, SUFFIX)(ARG1, addr,
+ MEMTXATTRS_UNSPECIFIED, NULL);
+}
+
+static inline uint64_t glue(ldq_le_phys, SUFFIX)(ARG1_DECL, hwaddr addr)
+{
+ return glue(address_space_ldq_le, SUFFIX)(ARG1, addr,
+ MEMTXATTRS_UNSPECIFIED, NULL);
+}
+
+static inline uint64_t glue(ldq_be_phys, SUFFIX)(ARG1_DECL, hwaddr addr)
+{
+ return glue(address_space_ldq_be, SUFFIX)(ARG1, addr,
+ MEMTXATTRS_UNSPECIFIED, NULL);
+}
+
+static inline uint32_t glue(ldub_phys, SUFFIX)(ARG1_DECL, hwaddr addr)
+{
+ return glue(address_space_ldub, SUFFIX)(ARG1, addr,
+ MEMTXATTRS_UNSPECIFIED, NULL);
+}
+
+static inline uint32_t glue(lduw_le_phys, SUFFIX)(ARG1_DECL, hwaddr addr)
+{
+ return glue(address_space_lduw_le, SUFFIX)(ARG1, addr,
+ MEMTXATTRS_UNSPECIFIED, NULL);
+}
+
+static inline uint32_t glue(lduw_be_phys, SUFFIX)(ARG1_DECL, hwaddr addr)
+{
+ return glue(address_space_lduw_be, SUFFIX)(ARG1, addr,
+ MEMTXATTRS_UNSPECIFIED, NULL);
+}
+
+static inline void glue(stl_le_phys, SUFFIX)(ARG1_DECL, hwaddr addr, uint32_t val)
+{
+ glue(address_space_stl_le, SUFFIX)(ARG1, addr, val,
+ MEMTXATTRS_UNSPECIFIED, NULL);
+}
+
+static inline void glue(stl_be_phys, SUFFIX)(ARG1_DECL, hwaddr addr, uint32_t val)
+{
+ glue(address_space_stl_be, SUFFIX)(ARG1, addr, val,
+ MEMTXATTRS_UNSPECIFIED, NULL);
+}
+
+static inline void glue(stb_phys, SUFFIX)(ARG1_DECL, hwaddr addr, uint32_t val)
+{
+ glue(address_space_stb, SUFFIX)(ARG1, addr, val,
+ MEMTXATTRS_UNSPECIFIED, NULL);
+}
+
+static inline void glue(stw_le_phys, SUFFIX)(ARG1_DECL, hwaddr addr, uint32_t val)
+{
+ glue(address_space_stw_le, SUFFIX)(ARG1, addr, val,
+ MEMTXATTRS_UNSPECIFIED, NULL);
+}
+
+static inline void glue(stw_be_phys, SUFFIX)(ARG1_DECL, hwaddr addr, uint32_t val)
+{
+ glue(address_space_stw_be, SUFFIX)(ARG1, addr, val,
+ MEMTXATTRS_UNSPECIFIED, NULL);
+}
+
+static inline void glue(stq_le_phys, SUFFIX)(ARG1_DECL, hwaddr addr, uint64_t val)
+{
+ glue(address_space_stq_le, SUFFIX)(ARG1, addr, val,
+ MEMTXATTRS_UNSPECIFIED, NULL);
+}
+
+static inline void glue(stq_be_phys, SUFFIX)(ARG1_DECL, hwaddr addr, uint64_t val)
+{
+ glue(address_space_stq_be, SUFFIX)(ARG1, addr, val,
+ MEMTXATTRS_UNSPECIFIED, NULL);
+}
+#endif
+
+#undef ARG1_DECL
+#undef ARG1
+#undef SUFFIX
+#undef TARGET_ENDIANNESS
diff --git a/memory_ldst.inc.c b/memory_ldst.inc.c
index 5dbff9c..25d6125 100644
--- a/memory_ldst.inc.c
+++ b/memory_ldst.inc.c
@@ -95,24 +95,6 @@ uint32_t glue(address_space_ldl_be, SUFFIX)(ARG1_DECL,
DEVICE_BIG_ENDIAN);
}
-uint32_t glue(ldl_phys, SUFFIX)(ARG1_DECL, hwaddr addr)
-{
- return glue(address_space_ldl, SUFFIX)(ARG1, addr,
- MEMTXATTRS_UNSPECIFIED, NULL);
-}
-
-uint32_t glue(ldl_le_phys, SUFFIX)(ARG1_DECL, hwaddr addr)
-{
- return glue(address_space_ldl_le, SUFFIX)(ARG1, addr,
- MEMTXATTRS_UNSPECIFIED, NULL);
-}
-
-uint32_t glue(ldl_be_phys, SUFFIX)(ARG1_DECL, hwaddr addr)
-{
- return glue(address_space_ldl_be, SUFFIX)(ARG1, addr,
- MEMTXATTRS_UNSPECIFIED, NULL);
-}
-
/* warning: addr must be aligned */
static inline uint64_t glue(address_space_ldq_internal, SUFFIX)(ARG1_DECL,
hwaddr addr, MemTxAttrs attrs, MemTxResult *result,
@@ -189,24 +171,6 @@ uint64_t glue(address_space_ldq_be, SUFFIX)(ARG1_DECL,
DEVICE_BIG_ENDIAN);
}
-uint64_t glue(ldq_phys, SUFFIX)(ARG1_DECL, hwaddr addr)
-{
- return glue(address_space_ldq, SUFFIX)(ARG1, addr,
- MEMTXATTRS_UNSPECIFIED, NULL);
-}
-
-uint64_t glue(ldq_le_phys, SUFFIX)(ARG1_DECL, hwaddr addr)
-{
- return glue(address_space_ldq_le, SUFFIX)(ARG1, addr,
- MEMTXATTRS_UNSPECIFIED, NULL);
-}
-
-uint64_t glue(ldq_be_phys, SUFFIX)(ARG1_DECL, hwaddr addr)
-{
- return glue(address_space_ldq_be, SUFFIX)(ARG1, addr,
- MEMTXATTRS_UNSPECIFIED, NULL);
-}
-
uint32_t glue(address_space_ldub, SUFFIX)(ARG1_DECL,
hwaddr addr, MemTxAttrs attrs, MemTxResult *result)
{
@@ -241,12 +205,6 @@ uint32_t glue(address_space_ldub, SUFFIX)(ARG1_DECL,
return val;
}
-uint32_t glue(ldub_phys, SUFFIX)(ARG1_DECL, hwaddr addr)
-{
- return glue(address_space_ldub, SUFFIX)(ARG1, addr,
- MEMTXATTRS_UNSPECIFIED, NULL);
-}
-
/* warning: addr must be aligned */
static inline uint32_t glue(address_space_lduw_internal, SUFFIX)(ARG1_DECL,
hwaddr addr, MemTxAttrs attrs, MemTxResult *result,
@@ -323,24 +281,6 @@ uint32_t glue(address_space_lduw_be, SUFFIX)(ARG1_DECL,
DEVICE_BIG_ENDIAN);
}
-uint32_t glue(lduw_phys, SUFFIX)(ARG1_DECL, hwaddr addr)
-{
- return glue(address_space_lduw, SUFFIX)(ARG1, addr,
- MEMTXATTRS_UNSPECIFIED, NULL);
-}
-
-uint32_t glue(lduw_le_phys, SUFFIX)(ARG1_DECL, hwaddr addr)
-{
- return glue(address_space_lduw_le, SUFFIX)(ARG1, addr,
- MEMTXATTRS_UNSPECIFIED, NULL);
-}
-
-uint32_t glue(lduw_be_phys, SUFFIX)(ARG1_DECL, hwaddr addr)
-{
- return glue(address_space_lduw_be, SUFFIX)(ARG1, addr,
- MEMTXATTRS_UNSPECIFIED, NULL);
-}
-
/* warning: addr must be aligned. The ram page is not masked as dirty
and the code inside is not invalidated. It is useful if the dirty
bits are used to track modified PTEs */
@@ -380,12 +320,6 @@ void glue(address_space_stl_notdirty, SUFFIX)(ARG1_DECL,
RCU_READ_UNLOCK();
}
-void glue(stl_phys_notdirty, SUFFIX)(ARG1_DECL, hwaddr addr, uint32_t val)
-{
- glue(address_space_stl_notdirty, SUFFIX)(ARG1, addr, val,
- MEMTXATTRS_UNSPECIFIED, NULL);
-}
-
/* warning: addr must be aligned */
static inline void glue(address_space_stl_internal, SUFFIX)(ARG1_DECL,
hwaddr addr, uint32_t val, MemTxAttrs attrs,
@@ -460,24 +394,6 @@ void glue(address_space_stl_be, SUFFIX)(ARG1_DECL,
result, DEVICE_BIG_ENDIAN);
}
-void glue(stl_phys, SUFFIX)(ARG1_DECL, hwaddr addr, uint32_t val)
-{
- glue(address_space_stl, SUFFIX)(ARG1, addr, val,
- MEMTXATTRS_UNSPECIFIED, NULL);
-}
-
-void glue(stl_le_phys, SUFFIX)(ARG1_DECL, hwaddr addr, uint32_t val)
-{
- glue(address_space_stl_le, SUFFIX)(ARG1, addr, val,
- MEMTXATTRS_UNSPECIFIED, NULL);
-}
-
-void glue(stl_be_phys, SUFFIX)(ARG1_DECL, hwaddr addr, uint32_t val)
-{
- glue(address_space_stl_be, SUFFIX)(ARG1, addr, val,
- MEMTXATTRS_UNSPECIFIED, NULL);
-}
-
void glue(address_space_stb, SUFFIX)(ARG1_DECL,
hwaddr addr, uint32_t val, MemTxAttrs attrs, MemTxResult *result)
{
@@ -509,12 +425,6 @@ void glue(address_space_stb, SUFFIX)(ARG1_DECL,
RCU_READ_UNLOCK();
}
-void glue(stb_phys, SUFFIX)(ARG1_DECL, hwaddr addr, uint32_t val)
-{
- glue(address_space_stb, SUFFIX)(ARG1, addr, val,
- MEMTXATTRS_UNSPECIFIED, NULL);
-}
-
/* warning: addr must be aligned */
static inline void glue(address_space_stw_internal, SUFFIX)(ARG1_DECL,
hwaddr addr, uint32_t val, MemTxAttrs attrs,
@@ -589,24 +499,6 @@ void glue(address_space_stw_be, SUFFIX)(ARG1_DECL,
DEVICE_BIG_ENDIAN);
}
-void glue(stw_phys, SUFFIX)(ARG1_DECL, hwaddr addr, uint32_t val)
-{
- glue(address_space_stw, SUFFIX)(ARG1, addr, val,
- MEMTXATTRS_UNSPECIFIED, NULL);
-}
-
-void glue(stw_le_phys, SUFFIX)(ARG1_DECL, hwaddr addr, uint32_t val)
-{
- glue(address_space_stw_le, SUFFIX)(ARG1, addr, val,
- MEMTXATTRS_UNSPECIFIED, NULL);
-}
-
-void glue(stw_be_phys, SUFFIX)(ARG1_DECL, hwaddr addr, uint32_t val)
-{
- glue(address_space_stw_be, SUFFIX)(ARG1, addr, val,
- MEMTXATTRS_UNSPECIFIED, NULL);
-}
-
static void glue(address_space_stq_internal, SUFFIX)(ARG1_DECL,
hwaddr addr, uint64_t val, MemTxAttrs attrs,
MemTxResult *result, enum device_endian endian)
@@ -680,24 +572,6 @@ void glue(address_space_stq_be, SUFFIX)(ARG1_DECL,
DEVICE_BIG_ENDIAN);
}
-void glue(stq_phys, SUFFIX)(ARG1_DECL, hwaddr addr, uint64_t val)
-{
- glue(address_space_stq, SUFFIX)(ARG1, addr, val,
- MEMTXATTRS_UNSPECIFIED, NULL);
-}
-
-void glue(stq_le_phys, SUFFIX)(ARG1_DECL, hwaddr addr, uint64_t val)
-{
- glue(address_space_stq_le, SUFFIX)(ARG1, addr, val,
- MEMTXATTRS_UNSPECIFIED, NULL);
-}
-
-void glue(stq_be_phys, SUFFIX)(ARG1_DECL, hwaddr addr, uint64_t val)
-{
- glue(address_space_stq_be, SUFFIX)(ARG1, addr, val,
- MEMTXATTRS_UNSPECIFIED, NULL);
-}
-
#undef ARG1_DECL
#undef ARG1
#undef SUFFIX
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* [Qemu-devel] [PULL 08/30] exec: small changes to flatview_do_translate
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
` (6 preceding siblings ...)
2018-05-08 22:14 ` [Qemu-devel] [PULL 07/30] exec: move memory access declarations to a common header, inline *_phys functions Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 09/30] exec: extract address_space_translate_iommu, fix page_mask corner case Paolo Bonzini
` (22 subsequent siblings)
30 siblings, 0 replies; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel
Prepare for extracting the IOMMU part to a separate function. Mostly
cosmetic; the only semantic change is that, if there is more than one
cascaded IOMMU and the second one fails to translate, *plen_out is now
adjusted according to the page mask of the first IOMMU.
Reviewed-by: Peter Xu <peterx@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
exec.c | 33 ++++++++++++++-------------------
1 file changed, 14 insertions(+), 19 deletions(-)
diff --git a/exec.c b/exec.c
index c7fcefa..8354cdc 100644
--- a/exec.c
+++ b/exec.c
@@ -476,6 +476,7 @@ address_space_translate_internal(AddressSpaceDispatch *d, hwaddr addr, hwaddr *x
* would tell. It can be @NULL if we don't care about it.
* @is_write: whether the translation operation is for write
* @is_mmio: whether this can be MMIO, set true if it can
+ * @target_as: the address space targeted by the IOMMU
*
* This function is called from RCU critical section
*/
@@ -495,14 +496,14 @@ static MemoryRegionSection flatview_do_translate(FlatView *fv,
hwaddr page_mask = (hwaddr)(-1);
hwaddr plen = (hwaddr)(-1);
- if (plen_out) {
- plen = *plen_out;
+ if (!plen_out) {
+ plen_out = &plen;
}
for (;;) {
section = address_space_translate_internal(
- flatview_to_dispatch(fv), addr, &addr,
- &plen, is_mmio);
+ flatview_to_dispatch(fv), addr, xlat,
+ plen_out, is_mmio);
iommu_mr = memory_region_get_iommu(section->mr);
if (!iommu_mr) {
@@ -510,35 +511,29 @@ static MemoryRegionSection flatview_do_translate(FlatView *fv,
}
imrc = memory_region_get_iommu_class_nocheck(iommu_mr);
+ addr = *xlat;
iotlb = imrc->translate(iommu_mr, addr, is_write ?
IOMMU_WO : IOMMU_RO);
- addr = ((iotlb.translated_addr & ~iotlb.addr_mask)
- | (addr & iotlb.addr_mask));
- page_mask &= iotlb.addr_mask;
- plen = MIN(plen, (addr | iotlb.addr_mask) - addr + 1);
if (!(iotlb.perm & (1 << is_write))) {
goto translate_fail;
}
+ addr = ((iotlb.translated_addr & ~iotlb.addr_mask)
+ | (addr & iotlb.addr_mask));
+ page_mask &= iotlb.addr_mask;
+ *plen_out = MIN(*plen_out, (addr | iotlb.addr_mask) - addr + 1);
fv = address_space_to_flatview(iotlb.target_as);
*target_as = iotlb.target_as;
}
- *xlat = addr;
-
- if (page_mask == (hwaddr)(-1)) {
- /* Not behind an IOMMU, use default page size. */
- page_mask = ~TARGET_PAGE_MASK;
- }
-
if (page_mask_out) {
+ if (page_mask == (hwaddr)(-1)) {
+ /* Not behind an IOMMU, use default page size. */
+ page_mask = ~TARGET_PAGE_MASK;
+ }
*page_mask_out = page_mask;
}
- if (plen_out) {
- *plen_out = plen;
- }
-
return *section;
translate_fail:
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* [Qemu-devel] [PULL 09/30] exec: extract address_space_translate_iommu, fix page_mask corner case
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
` (7 preceding siblings ...)
2018-05-08 22:14 ` [Qemu-devel] [PULL 08/30] exec: small changes to flatview_do_translate Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 10/30] exec: reintroduce MemoryRegion caching Paolo Bonzini
` (21 subsequent siblings)
30 siblings, 0 replies; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel
This will be used to process IOMMUs in a MemoryRegionCache. This
includes a small bugfix, in that the returned page_mask is now
correctly -1 if the IOMMU memory region maps the entire address
space directly. Previously, address_space_get_iotlb_entry would
return ~TARGET_PAGE_MASK.
Reviewed-by: Peter Xu <peterx@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
exec.c | 110 ++++++++++++++++++++++++++++++++++++++++++++---------------------
1 file changed, 75 insertions(+), 35 deletions(-)
diff --git a/exec.c b/exec.c
index 8354cdc..5f98106 100644
--- a/exec.c
+++ b/exec.c
@@ -462,6 +462,70 @@ address_space_translate_internal(AddressSpaceDispatch *d, hwaddr addr, hwaddr *x
}
/**
+ * address_space_translate_iommu - translate an address through an IOMMU
+ * memory region and then through the target address space.
+ *
+ * @iommu_mr: the IOMMU memory region that we start the translation from
+ * @addr: the address to be translated through the MMU
+ * @xlat: the translated address offset within the destination memory region.
+ * It cannot be %NULL.
+ * @plen_out: valid read/write length of the translated address. It
+ * cannot be %NULL.
+ * @page_mask_out: page mask for the translated address. This
+ * should only be meaningful for IOMMU translated
+ * addresses, since there may be huge pages that this bit
+ * would tell. It can be %NULL if we don't care about it.
+ * @is_write: whether the translation operation is for write
+ * @is_mmio: whether this can be MMIO, set true if it can
+ * @target_as: the address space targeted by the IOMMU
+ *
+ * This function is called from RCU critical section. It is the common
+ * part of flatview_do_translate and address_space_translate_cached.
+ */
+static MemoryRegionSection address_space_translate_iommu(IOMMUMemoryRegion *iommu_mr,
+ hwaddr *xlat,
+ hwaddr *plen_out,
+ hwaddr *page_mask_out,
+ bool is_write,
+ bool is_mmio,
+ AddressSpace **target_as)
+{
+ MemoryRegionSection *section;
+ hwaddr page_mask = (hwaddr)-1;
+
+ do {
+ hwaddr addr = *xlat;
+ IOMMUMemoryRegionClass *imrc = memory_region_get_iommu_class_nocheck(iommu_mr);
+ IOMMUTLBEntry iotlb = imrc->translate(iommu_mr, addr, is_write ?
+ IOMMU_WO : IOMMU_RO);
+
+ if (!(iotlb.perm & (1 << is_write))) {
+ goto unassigned;
+ }
+
+ addr = ((iotlb.translated_addr & ~iotlb.addr_mask)
+ | (addr & iotlb.addr_mask));
+ page_mask &= iotlb.addr_mask;
+ *plen_out = MIN(*plen_out, (addr | iotlb.addr_mask) - addr + 1);
+ *target_as = iotlb.target_as;
+
+ section = address_space_translate_internal(
+ address_space_to_dispatch(iotlb.target_as), addr, xlat,
+ plen_out, is_mmio);
+
+ iommu_mr = memory_region_get_iommu(section->mr);
+ } while (unlikely(iommu_mr));
+
+ if (page_mask_out) {
+ *page_mask_out = page_mask;
+ }
+ return *section;
+
+unassigned:
+ return (MemoryRegionSection) { .mr = &io_mem_unassigned };
+}
+
+/**
* flatview_do_translate - translate an address in FlatView
*
* @fv: the flat view that we want to translate on
@@ -489,55 +553,31 @@ static MemoryRegionSection flatview_do_translate(FlatView *fv,
bool is_mmio,
AddressSpace **target_as)
{
- IOMMUTLBEntry iotlb;
MemoryRegionSection *section;
IOMMUMemoryRegion *iommu_mr;
- IOMMUMemoryRegionClass *imrc;
- hwaddr page_mask = (hwaddr)(-1);
hwaddr plen = (hwaddr)(-1);
if (!plen_out) {
plen_out = &plen;
}
- for (;;) {
- section = address_space_translate_internal(
- flatview_to_dispatch(fv), addr, xlat,
- plen_out, is_mmio);
-
- iommu_mr = memory_region_get_iommu(section->mr);
- if (!iommu_mr) {
- break;
- }
- imrc = memory_region_get_iommu_class_nocheck(iommu_mr);
-
- addr = *xlat;
- iotlb = imrc->translate(iommu_mr, addr, is_write ?
- IOMMU_WO : IOMMU_RO);
- if (!(iotlb.perm & (1 << is_write))) {
- goto translate_fail;
- }
+ section = address_space_translate_internal(
+ flatview_to_dispatch(fv), addr, xlat,
+ plen_out, is_mmio);
- addr = ((iotlb.translated_addr & ~iotlb.addr_mask)
- | (addr & iotlb.addr_mask));
- page_mask &= iotlb.addr_mask;
- *plen_out = MIN(*plen_out, (addr | iotlb.addr_mask) - addr + 1);
- fv = address_space_to_flatview(iotlb.target_as);
- *target_as = iotlb.target_as;
+ iommu_mr = memory_region_get_iommu(section->mr);
+ if (unlikely(iommu_mr)) {
+ return address_space_translate_iommu(iommu_mr, xlat,
+ plen_out, page_mask_out,
+ is_write, is_mmio,
+ target_as);
}
-
if (page_mask_out) {
- if (page_mask == (hwaddr)(-1)) {
- /* Not behind an IOMMU, use default page size. */
- page_mask = ~TARGET_PAGE_MASK;
- }
- *page_mask_out = page_mask;
+ /* Not behind an IOMMU, use default page size. */
+ *page_mask_out = ~TARGET_PAGE_MASK;
}
return *section;
-
-translate_fail:
- return (MemoryRegionSection) { .mr = &io_mem_unassigned };
}
/* Called from RCU critical section */
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* [Qemu-devel] [PULL 10/30] exec: reintroduce MemoryRegion caching
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
` (8 preceding siblings ...)
2018-05-08 22:14 ` [Qemu-devel] [PULL 09/30] exec: extract address_space_translate_iommu, fix page_mask corner case Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 11/30] qemu-thread: always keep the posix wrapper layer Paolo Bonzini
` (20 subsequent siblings)
30 siblings, 0 replies; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel
MemoryRegionCache was reverted to "normal" address_space_* operations
for 2.9, due to lack of support for IOMMUs. Reinstate the
optimizations, caching only the IOMMU translation at address_cache_init
but not the IOMMU lookup and target AddressSpace translation are not
cached; now that MemoryRegionCache supports IOMMUs, it becomes more widely
applicable too.
The inlined fast path is defined in memory_ldst_cached.inc.h, while the
slow path uses memory_ldst.inc.c as before. The smaller fast path causes
a little code size reduction in MemoryRegionCache users:
hw/virtio/virtio.o text size before: 32373
hw/virtio/virtio.o text size after: 31941
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
exec.c | 121 ++++++++++++++++++++++++++++++----
include/exec/cpu-all.h | 6 +-
include/exec/memory-internal.h | 3 +
include/exec/memory.h | 58 ++++++++++++++--
include/exec/memory_ldst_cached.inc.h | 108 ++++++++++++++++++++++++++++++
memory.c | 4 +-
6 files changed, 280 insertions(+), 20 deletions(-)
create mode 100644 include/exec/memory_ldst_cached.inc.h
diff --git a/exec.c b/exec.c
index 5f98106..ffa1099 100644
--- a/exec.c
+++ b/exec.c
@@ -3641,33 +3641,130 @@ int64_t address_space_cache_init(MemoryRegionCache *cache,
hwaddr len,
bool is_write)
{
- cache->len = len;
- cache->as = as;
- cache->xlat = addr;
- return len;
+ AddressSpaceDispatch *d;
+ hwaddr l;
+ MemoryRegion *mr;
+
+ assert(len > 0);
+
+ l = len;
+ cache->fv = address_space_get_flatview(as);
+ d = flatview_to_dispatch(cache->fv);
+ cache->mrs = *address_space_translate_internal(d, addr, &cache->xlat, &l, true);
+
+ mr = cache->mrs.mr;
+ memory_region_ref(mr);
+ if (memory_access_is_direct(mr, is_write)) {
+ l = flatview_extend_translation(cache->fv, addr, len, mr,
+ cache->xlat, l, is_write);
+ cache->ptr = qemu_ram_ptr_length(mr->ram_block, cache->xlat, &l, true);
+ } else {
+ cache->ptr = NULL;
+ }
+
+ cache->len = l;
+ cache->is_write = is_write;
+ return l;
}
void address_space_cache_invalidate(MemoryRegionCache *cache,
hwaddr addr,
hwaddr access_len)
{
+ assert(cache->is_write);
+ if (likely(cache->ptr)) {
+ invalidate_and_set_dirty(cache->mrs.mr, addr + cache->xlat, access_len);
+ }
}
void address_space_cache_destroy(MemoryRegionCache *cache)
{
- cache->as = NULL;
+ if (!cache->mrs.mr) {
+ return;
+ }
+
+ if (xen_enabled()) {
+ xen_invalidate_map_cache_entry(cache->ptr);
+ }
+ memory_region_unref(cache->mrs.mr);
+ flatview_unref(cache->fv);
+ cache->mrs.mr = NULL;
+ cache->fv = NULL;
+}
+
+/* Called from RCU critical section. This function has the same
+ * semantics as address_space_translate, but it only works on a
+ * predefined range of a MemoryRegion that was mapped with
+ * address_space_cache_init.
+ */
+static inline MemoryRegion *address_space_translate_cached(
+ MemoryRegionCache *cache, hwaddr addr, hwaddr *xlat,
+ hwaddr *plen, bool is_write)
+{
+ MemoryRegionSection section;
+ MemoryRegion *mr;
+ IOMMUMemoryRegion *iommu_mr;
+ AddressSpace *target_as;
+
+ assert(!cache->ptr);
+ *xlat = addr + cache->xlat;
+
+ mr = cache->mrs.mr;
+ iommu_mr = memory_region_get_iommu(mr);
+ if (!iommu_mr) {
+ /* MMIO region. */
+ return mr;
+ }
+
+ section = address_space_translate_iommu(iommu_mr, xlat, plen,
+ NULL, is_write, true,
+ &target_as);
+ return section.mr;
+}
+
+/* Called from RCU critical section. address_space_read_cached uses this
+ * out of line function when the target is an MMIO or IOMMU region.
+ */
+void
+address_space_read_cached_slow(MemoryRegionCache *cache, hwaddr addr,
+ void *buf, int len)
+{
+ hwaddr addr1, l;
+ MemoryRegion *mr;
+
+ l = len;
+ mr = address_space_translate_cached(cache, addr, &addr1, &l, false);
+ flatview_read_continue(cache->fv,
+ addr, MEMTXATTRS_UNSPECIFIED, buf, len,
+ addr1, l, mr);
+}
+
+/* Called from RCU critical section. address_space_write_cached uses this
+ * out of line function when the target is an MMIO or IOMMU region.
+ */
+void
+address_space_write_cached_slow(MemoryRegionCache *cache, hwaddr addr,
+ const void *buf, int len)
+{
+ hwaddr addr1, l;
+ MemoryRegion *mr;
+
+ l = len;
+ mr = address_space_translate_cached(cache, addr, &addr1, &l, true);
+ flatview_write_continue(cache->fv,
+ addr, MEMTXATTRS_UNSPECIFIED, buf, len,
+ addr1, l, mr);
}
#define ARG1_DECL MemoryRegionCache *cache
#define ARG1 cache
-#define SUFFIX _cached
-#define TRANSLATE(addr, ...) \
- address_space_translate(cache->as, cache->xlat + (addr), __VA_ARGS__)
-#define IS_DIRECT(mr, is_write) true
-#define MAP_RAM(mr, ofs) qemu_map_ram_ptr((mr)->ram_block, ofs)
+#define SUFFIX _cached_slow
+#define TRANSLATE(...) address_space_translate_cached(cache, __VA_ARGS__)
+#define IS_DIRECT(mr, is_write) memory_access_is_direct(mr, is_write)
+#define MAP_RAM(mr, ofs) (cache->ptr + (ofs - cache->xlat))
#define INVALIDATE(mr, ofs, len) invalidate_and_set_dirty(mr, ofs, len)
-#define RCU_READ_LOCK() rcu_read_lock()
-#define RCU_READ_UNLOCK() rcu_read_unlock()
+#define RCU_READ_LOCK() ((void)0)
+#define RCU_READ_UNLOCK() ((void)0)
#include "memory_ldst.inc.c"
/* virtual memory access for debug (includes writing to ROM) */
diff --git a/include/exec/cpu-all.h b/include/exec/cpu-all.h
index 173edd1..a635f53 100644
--- a/include/exec/cpu-all.h
+++ b/include/exec/cpu-all.h
@@ -175,7 +175,7 @@ extern unsigned long reserved_va;
#define TARGET_ENDIANNESS
#include "exec/memory_ldst.inc.h"
-#define SUFFIX _cached
+#define SUFFIX _cached_slow
#define ARG1 cache
#define ARG1_DECL MemoryRegionCache *cache
#define TARGET_ENDIANNESS
@@ -193,6 +193,10 @@ static inline void stl_phys_notdirty(AddressSpace *as, hwaddr addr, uint32_t val
#define TARGET_ENDIANNESS
#include "exec/memory_ldst_phys.inc.h"
+/* Inline fast path for direct RAM access. */
+#define ENDIANNESS
+#include "exec/memory_ldst_cached.inc.h"
+
#define SUFFIX _cached
#define ARG1 cache
#define ARG1_DECL MemoryRegionCache *cache
diff --git a/include/exec/memory-internal.h b/include/exec/memory-internal.h
index 6a5ee42..58399b9 100644
--- a/include/exec/memory-internal.h
+++ b/include/exec/memory-internal.h
@@ -31,6 +31,9 @@ static inline AddressSpaceDispatch *address_space_to_dispatch(AddressSpace *as)
return flatview_to_dispatch(address_space_to_flatview(as));
}
+FlatView *address_space_get_flatview(AddressSpace *as);
+void flatview_unref(FlatView *view);
+
extern const MemoryRegionOps unassigned_mem_ops;
bool memory_region_access_valid(MemoryRegion *mr, hwaddr addr,
diff --git a/include/exec/memory.h b/include/exec/memory.h
index ca361bc..525619a 100644
--- a/include/exec/memory.h
+++ b/include/exec/memory.h
@@ -1688,12 +1688,16 @@ MemTxResult address_space_write(AddressSpace *as, hwaddr addr,
#include "exec/memory_ldst_phys.inc.h"
struct MemoryRegionCache {
+ void *ptr;
hwaddr xlat;
hwaddr len;
- AddressSpace *as;
+ FlatView *fv;
+ MemoryRegionSection mrs;
+ bool is_write;
};
-#define MEMORY_REGION_CACHE_INVALID ((MemoryRegionCache) { .as = NULL })
+#define MEMORY_REGION_CACHE_INVALID ((MemoryRegionCache) { .mrs.mr = NULL })
+
/* address_space_ld*_cached: load from a cached #MemoryRegion
* address_space_st*_cached: store into a cached #MemoryRegion
@@ -1719,11 +1723,40 @@ struct MemoryRegionCache {
* if NULL, this information is discarded
*/
-#define SUFFIX _cached
+#define SUFFIX _cached_slow
#define ARG1 cache
#define ARG1_DECL MemoryRegionCache *cache
#include "exec/memory_ldst.inc.h"
+/* Inline fast path for direct RAM access. */
+static inline uint8_t address_space_ldub_cached(MemoryRegionCache *cache,
+ hwaddr addr, MemTxAttrs attrs, MemTxResult *result)
+{
+ assert(addr < cache->len);
+ if (likely(cache->ptr)) {
+ return ldub_p(cache->ptr + addr);
+ } else {
+ return address_space_ldub_cached_slow(cache, addr, attrs, result);
+ }
+}
+
+static inline void address_space_stb_cached(MemoryRegionCache *cache,
+ hwaddr addr, uint32_t val, MemTxAttrs attrs, MemTxResult *result)
+{
+ assert(addr < cache->len);
+ if (likely(cache->ptr)) {
+ stb_p(cache->ptr + addr, val);
+ } else {
+ address_space_stb_cached_slow(cache, addr, val, attrs, result);
+ }
+}
+
+#define ENDIANNESS _le
+#include "exec/memory_ldst_cached.inc.h"
+
+#define ENDIANNESS _be
+#include "exec/memory_ldst_cached.inc.h"
+
#define SUFFIX _cached
#define ARG1 cache
#define ARG1_DECL MemoryRegionCache *cache
@@ -1860,6 +1893,13 @@ MemTxResult flatview_read_continue(FlatView *fv, hwaddr addr,
MemoryRegion *mr);
void *qemu_map_ram_ptr(RAMBlock *ram_block, ram_addr_t addr);
+/* Internal functions, part of the implementation of address_space_read_cached
+ * and address_space_write_cached. */
+void address_space_read_cached_slow(MemoryRegionCache *cache,
+ hwaddr addr, void *buf, int len);
+void address_space_write_cached_slow(MemoryRegionCache *cache,
+ hwaddr addr, const void *buf, int len);
+
static inline bool memory_access_is_direct(MemoryRegion *mr, bool is_write)
{
if (is_write) {
@@ -1928,7 +1968,11 @@ address_space_read_cached(MemoryRegionCache *cache, hwaddr addr,
void *buf, int len)
{
assert(addr < cache->len && len <= cache->len - addr);
- address_space_read(cache->as, cache->xlat + addr, MEMTXATTRS_UNSPECIFIED, buf, len);
+ if (likely(cache->ptr)) {
+ memcpy(buf, cache->ptr + addr, len);
+ } else {
+ address_space_read_cached_slow(cache, addr, buf, len);
+ }
}
/**
@@ -1944,7 +1988,11 @@ address_space_write_cached(MemoryRegionCache *cache, hwaddr addr,
void *buf, int len)
{
assert(addr < cache->len && len <= cache->len - addr);
- address_space_write(cache->as, cache->xlat + addr, MEMTXATTRS_UNSPECIFIED, buf, len);
+ if (likely(cache->ptr)) {
+ memcpy(cache->ptr + addr, buf, len);
+ } else {
+ address_space_write_cached_slow(cache, addr, buf, len);
+ }
}
#endif
diff --git a/include/exec/memory_ldst_cached.inc.h b/include/exec/memory_ldst_cached.inc.h
new file mode 100644
index 0000000..fd4bbb4
--- /dev/null
+++ b/include/exec/memory_ldst_cached.inc.h
@@ -0,0 +1,108 @@
+/*
+ * Memory access templates for MemoryRegionCache
+ *
+ * Copyright (c) 2018 Red Hat, Inc.
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library; if not, see <http://www.gnu.org/licenses/>.
+ */
+
+#define ADDRESS_SPACE_LD_CACHED(size) \
+ glue(glue(address_space_ld, size), glue(ENDIANNESS, _cached))
+#define ADDRESS_SPACE_LD_CACHED_SLOW(size) \
+ glue(glue(address_space_ld, size), glue(ENDIANNESS, _cached_slow))
+#define LD_P(size) \
+ glue(glue(ld, size), glue(ENDIANNESS, _p))
+
+static inline uint32_t ADDRESS_SPACE_LD_CACHED(l)(MemoryRegionCache *cache,
+ hwaddr addr, MemTxAttrs attrs, MemTxResult *result)
+{
+ assert(addr < cache->len && 4 <= cache->len - addr);
+ if (likely(cache->ptr)) {
+ return LD_P(l)(cache->ptr + addr);
+ } else {
+ return ADDRESS_SPACE_LD_CACHED_SLOW(l)(cache, addr, attrs, result);
+ }
+}
+
+static inline uint64_t ADDRESS_SPACE_LD_CACHED(q)(MemoryRegionCache *cache,
+ hwaddr addr, MemTxAttrs attrs, MemTxResult *result)
+{
+ assert(addr < cache->len && 8 <= cache->len - addr);
+ if (likely(cache->ptr)) {
+ return LD_P(q)(cache->ptr + addr);
+ } else {
+ return ADDRESS_SPACE_LD_CACHED_SLOW(q)(cache, addr, attrs, result);
+ }
+}
+
+static inline uint32_t ADDRESS_SPACE_LD_CACHED(uw)(MemoryRegionCache *cache,
+ hwaddr addr, MemTxAttrs attrs, MemTxResult *result)
+{
+ assert(addr < cache->len && 2 <= cache->len - addr);
+ if (likely(cache->ptr)) {
+ return LD_P(uw)(cache->ptr + addr);
+ } else {
+ return ADDRESS_SPACE_LD_CACHED_SLOW(uw)(cache, addr, attrs, result);
+ }
+}
+
+#undef ADDRESS_SPACE_LD_CACHED
+#undef ADDRESS_SPACE_LD_CACHED_SLOW
+#undef LD_P
+
+#define ADDRESS_SPACE_ST_CACHED(size) \
+ glue(glue(address_space_st, size), glue(ENDIANNESS, _cached))
+#define ADDRESS_SPACE_ST_CACHED_SLOW(size) \
+ glue(glue(address_space_st, size), glue(ENDIANNESS, _cached_slow))
+#define ST_P(size) \
+ glue(glue(st, size), glue(ENDIANNESS, _p))
+
+static inline void ADDRESS_SPACE_ST_CACHED(l)(MemoryRegionCache *cache,
+ hwaddr addr, uint32_t val, MemTxAttrs attrs, MemTxResult *result)
+{
+ assert(addr < cache->len && 4 <= cache->len - addr);
+ if (likely(cache->ptr)) {
+ ST_P(l)(cache->ptr + addr, val);
+ } else {
+ ADDRESS_SPACE_ST_CACHED_SLOW(l)(cache, addr, val, attrs, result);
+ }
+}
+
+static inline void ADDRESS_SPACE_ST_CACHED(w)(MemoryRegionCache *cache,
+ hwaddr addr, uint32_t val, MemTxAttrs attrs, MemTxResult *result)
+{
+ assert(addr < cache->len && 2 <= cache->len - addr);
+ if (likely(cache->ptr)) {
+ ST_P(w)(cache->ptr + addr, val);
+ } else {
+ ADDRESS_SPACE_ST_CACHED_SLOW(w)(cache, addr, val, attrs, result);
+ }
+}
+
+static inline void ADDRESS_SPACE_ST_CACHED(q)(MemoryRegionCache *cache,
+ hwaddr addr, uint64_t val, MemTxAttrs attrs, MemTxResult *result)
+{
+ assert(addr < cache->len && 8 <= cache->len - addr);
+ if (likely(cache->ptr)) {
+ ST_P(q)(cache->ptr + addr, val);
+ } else {
+ ADDRESS_SPACE_ST_CACHED_SLOW(q)(cache, addr, val, attrs, result);
+ }
+}
+
+#undef ADDRESS_SPACE_ST_CACHED
+#undef ADDRESS_SPACE_ST_CACHED_SLOW
+#undef ST_P
+
+#undef ENDIANNESS
diff --git a/memory.c b/memory.c
index e70b64b..fc7f9b7 100644
--- a/memory.c
+++ b/memory.c
@@ -298,7 +298,7 @@ static bool flatview_ref(FlatView *view)
return atomic_fetch_inc_nonzero(&view->ref) > 0;
}
-static void flatview_unref(FlatView *view)
+void flatview_unref(FlatView *view)
{
if (atomic_fetch_dec(&view->ref) == 1) {
trace_flatview_destroy_rcu(view, view->root);
@@ -822,7 +822,7 @@ static void address_space_add_del_ioeventfds(AddressSpace *as,
}
}
-static FlatView *address_space_get_flatview(AddressSpace *as)
+FlatView *address_space_get_flatview(AddressSpace *as)
{
FlatView *view;
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* [Qemu-devel] [PULL 11/30] qemu-thread: always keep the posix wrapper layer
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
` (9 preceding siblings ...)
2018-05-08 22:14 ` [Qemu-devel] [PULL 10/30] exec: reintroduce MemoryRegion caching Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 12/30] update-linux-headers: drop hyperv.h Paolo Bonzini
` (19 subsequent siblings)
30 siblings, 0 replies; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel; +Cc: Peter Xu
From: Peter Xu <peterx@redhat.com>
We will conditionally have a wrapper layer depending on whether the host
has the PTHREAD_SETNAME capability. It complicates stuff. Let's keep
the wrapper there; we opt out the pthread_setname_np() call only.
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20180412053444.17801-1-peterx@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
util/qemu-thread-posix.c | 33 +++++++++++++--------------------
1 file changed, 13 insertions(+), 20 deletions(-)
diff --git a/util/qemu-thread-posix.c b/util/qemu-thread-posix.c
index b789cf3..a1c34ba 100644
--- a/util/qemu-thread-posix.c
+++ b/util/qemu-thread-posix.c
@@ -482,7 +482,6 @@ static void __attribute__((constructor)) qemu_thread_atexit_init(void)
}
-#ifdef CONFIG_PTHREAD_SETNAME_NP
typedef struct {
void *(*start_routine)(void *);
void *arg;
@@ -495,16 +494,18 @@ static void *qemu_thread_start(void *args)
void *(*start_routine)(void *) = qemu_thread_args->start_routine;
void *arg = qemu_thread_args->arg;
+#ifdef CONFIG_PTHREAD_SETNAME_NP
/* Attempt to set the threads name; note that this is for debug, so
* we're not going to fail if we can't set it.
*/
- pthread_setname_np(pthread_self(), qemu_thread_args->name);
+ if (name_threads && qemu_thread_args->name) {
+ pthread_setname_np(pthread_self(), qemu_thread_args->name);
+ }
+#endif
g_free(qemu_thread_args->name);
g_free(qemu_thread_args);
return start_routine(arg);
}
-#endif
-
void qemu_thread_create(QemuThread *thread, const char *name,
void *(*start_routine)(void*),
@@ -513,6 +514,7 @@ void qemu_thread_create(QemuThread *thread, const char *name,
sigset_t set, oldset;
int err;
pthread_attr_t attr;
+ QemuThreadArgs *qemu_thread_args;
err = pthread_attr_init(&attr);
if (err) {
@@ -527,22 +529,13 @@ void qemu_thread_create(QemuThread *thread, const char *name,
sigfillset(&set);
pthread_sigmask(SIG_SETMASK, &set, &oldset);
-#ifdef CONFIG_PTHREAD_SETNAME_NP
- if (name_threads) {
- QemuThreadArgs *qemu_thread_args;
- qemu_thread_args = g_new0(QemuThreadArgs, 1);
- qemu_thread_args->name = g_strdup(name);
- qemu_thread_args->start_routine = start_routine;
- qemu_thread_args->arg = arg;
-
- err = pthread_create(&thread->thread, &attr,
- qemu_thread_start, qemu_thread_args);
- } else
-#endif
- {
- err = pthread_create(&thread->thread, &attr,
- start_routine, arg);
- }
+ qemu_thread_args = g_new0(QemuThreadArgs, 1);
+ qemu_thread_args->name = g_strdup(name);
+ qemu_thread_args->start_routine = start_routine;
+ qemu_thread_args->arg = arg;
+
+ err = pthread_create(&thread->thread, &attr,
+ qemu_thread_start, qemu_thread_args);
if (err)
error_exit(err, __func__);
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* [Qemu-devel] [PULL 12/30] update-linux-headers: drop hyperv.h
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
` (10 preceding siblings ...)
2018-05-08 22:14 ` [Qemu-devel] [PULL 11/30] qemu-thread: always keep the posix wrapper layer Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 13/30] accel: use g_strsplit for parsing accelerator names Paolo Bonzini
` (18 subsequent siblings)
30 siblings, 0 replies; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel; +Cc: Roman Kagan, Vitaly Kuznetsov
From: Roman Kagan <rkagan@virtuozzo.com>
As of mainline linux commit 5a485803221777013944cbd1a7cd5c62efba3ffa
"x86/hyper-v: move hyperv.h out of uapi" by Vitaly Kuznetsov, no linux
uapi header includes it, so we no longer need to create a stub for it.
Cc: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: Roman Kagan <rkagan@virtuozzo.com>
Message-Id: <20180413143354.17614-1-rkagan@virtuozzo.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
| 6 ------
1 file changed, 6 deletions(-)
--git a/scripts/update-linux-headers.sh b/scripts/update-linux-headers.sh
index 5b1d8dc..b65c03f 100755
--- a/scripts/update-linux-headers.sh
+++ b/scripts/update-linux-headers.sh
@@ -118,9 +118,6 @@ for arch in $ARCHLIST; do
cp "$tmpdir/include/asm/unistd-common.h" "$output/linux-headers/asm-arm/"
fi
if [ $arch = x86 ]; then
- cat <<-EOF >"$output/include/standard-headers/asm-x86/hyperv.h"
- /* this is a temporary placeholder until kvm_para.h stops including it */
-EOF
cp "$tmpdir/include/asm/unistd_32.h" "$output/linux-headers/asm-x86/"
cp "$tmpdir/include/asm/unistd_x32.h" "$output/linux-headers/asm-x86/"
cp "$tmpdir/include/asm/unistd_64.h" "$output/linux-headers/asm-x86/"
@@ -144,9 +141,6 @@ else
cp "$linux/COPYING" "$output/linux-headers"
fi
-cat <<EOF >$output/linux-headers/asm-x86/hyperv.h
-#include "standard-headers/asm-x86/hyperv.h"
-EOF
cat <<EOF >$output/linux-headers/linux/virtio_config.h
#include "standard-headers/linux/virtio_config.h"
EOF
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* [Qemu-devel] [PULL 13/30] accel: use g_strsplit for parsing accelerator names
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
` (11 preceding siblings ...)
2018-05-08 22:14 ` [Qemu-devel] [PULL 12/30] update-linux-headers: drop hyperv.h Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 14/30] opts: don't silently truncate long parameter keys Paolo Bonzini
` (17 subsequent siblings)
30 siblings, 0 replies; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel
From: Daniel P. Berrangé <berrange@redhat.com>
Instead of re-using the get_opt_name() method from QemuOpts to split a
string on ':', just use g_strsplit().
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Message-Id: <20180416111743.8473-2-berrange@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
---
accel/accel.c | 16 +++++++---------
include/qemu/option.h | 1 -
util/qemu-option.c | 3 ++-
3 files changed, 9 insertions(+), 11 deletions(-)
diff --git a/accel/accel.c b/accel/accel.c
index 9cfab11..966b2d8 100644
--- a/accel/accel.c
+++ b/accel/accel.c
@@ -70,8 +70,8 @@ static int accel_init_machine(AccelClass *acc, MachineState *ms)
void configure_accelerator(MachineState *ms)
{
- const char *accel, *p;
- char buf[10];
+ const char *accel;
+ char **accel_list, **tmp;
int ret;
bool accel_initialised = false;
bool init_failed = false;
@@ -83,13 +83,10 @@ void configure_accelerator(MachineState *ms)
accel = "tcg";
}
- p = accel;
- while (!accel_initialised && *p != '\0') {
- if (*p == ':') {
- p++;
- }
- p = get_opt_name(buf, sizeof(buf), p, ':');
- acc = accel_find(buf);
+ accel_list = g_strsplit(accel, ":", 0);
+
+ for (tmp = accel_list; !accel_initialised && tmp && *tmp; tmp++) {
+ acc = accel_find(*tmp);
if (!acc) {
continue;
}
@@ -107,6 +104,7 @@ void configure_accelerator(MachineState *ms)
accel_initialised = true;
}
}
+ g_strfreev(accel_list);
if (!accel_initialised) {
if (!init_failed) {
diff --git a/include/qemu/option.h b/include/qemu/option.h
index 306fdb5..1cfe5cb 100644
--- a/include/qemu/option.h
+++ b/include/qemu/option.h
@@ -28,7 +28,6 @@
#include "qemu/queue.h"
-const char *get_opt_name(char *buf, int buf_size, const char *p, char delim);
const char *get_opt_value(char *buf, int buf_size, const char *p);
void parse_option_size(const char *name, const char *value,
diff --git a/util/qemu-option.c b/util/qemu-option.c
index d0756fd..baca40f 100644
--- a/util/qemu-option.c
+++ b/util/qemu-option.c
@@ -49,7 +49,8 @@
* The return value is the position of the delimiter/zero byte after the option
* name in p.
*/
-const char *get_opt_name(char *buf, int buf_size, const char *p, char delim)
+static const char *get_opt_name(char *buf, int buf_size, const char *p,
+ char delim)
{
char *q;
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* [Qemu-devel] [PULL 14/30] opts: don't silently truncate long parameter keys
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
` (12 preceding siblings ...)
2018-05-08 22:14 ` [Qemu-devel] [PULL 13/30] accel: use g_strsplit for parsing accelerator names Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-09 5:46 ` Thomas Huth
2018-05-08 22:14 ` [Qemu-devel] [PULL 15/30] opts: don't silently truncate long option values Paolo Bonzini
` (16 subsequent siblings)
30 siblings, 1 reply; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel
From: Daniel P. Berrangé <berrange@redhat.com>
The existing QemuOpts parsing code uses a fixed size 128 byte buffer
for storing the parameter keys. If a key exceeded this size it was
silently truncate and no error reported to the user. This behaviour was
reasonable & harmless because traditionally the key names are all
statically declared, and it was known that no code was declaring a key
longer than 127 bytes. This assumption, however, ceased to be valid once
the block layer added support for dot-separate compound keys. This
syntax allows for keys that can be arbitrarily long, limited only by the
number of block drivers you can stack up. With this usage, silently
truncating the key name can never lead to correct behaviour.
Hopefully such truncation would turn into an error, when the block code
then tried to extract options later, but there's no guarantee that will
happen. It is conceivable that an option specified by the user may be
truncated and then ignored. This could have serious consequences,
possibly even leading to security problems if the ignored option set a
security relevant parameter.
If the operating system didn't limit the user's argv when spawning QEMU,
the code should honour whatever length arguments were given without
imposing its own length restrictions. This patch thus changes the code
to use a heap allocated buffer for storing the keys during parsing,
lifting the arbitrary length restriction.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Message-Id: <20180416111743.8473-3-berrange@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
---
tests/test-qemu-opts.c | 18 ------------------
util/qemu-option.c | 44 ++++++++++++++++++++++----------------------
2 files changed, 22 insertions(+), 40 deletions(-)
diff --git a/tests/test-qemu-opts.c b/tests/test-qemu-opts.c
index 77dd72b..7092e21 100644
--- a/tests/test-qemu-opts.c
+++ b/tests/test-qemu-opts.c
@@ -459,8 +459,6 @@ static void test_opts_parse(void)
{
Error *err = NULL;
QemuOpts *opts;
- char long_key[129];
- char *params;
/* Nothing */
opts = qemu_opts_parse(&opts_list_03, "", false, &error_abort);
@@ -471,22 +469,6 @@ static void test_opts_parse(void)
g_assert_cmpuint(opts_count(opts), ==, 1);
g_assert_cmpstr(qemu_opt_get(opts, ""), ==, "val");
- /* Long key */
- memset(long_key, 'a', 127);
- long_key[127] = 'z';
- long_key[128] = 0;
- params = g_strdup_printf("%s=v", long_key);
- opts = qemu_opts_parse(&opts_list_03, params + 1, NULL, &error_abort);
- g_assert_cmpuint(opts_count(opts), ==, 1);
- g_assert_cmpstr(qemu_opt_get(opts, long_key + 1), ==, "v");
-
- /* Overlong key gets truncated */
- opts = qemu_opts_parse(&opts_list_03, params, NULL, &error_abort);
- g_assert(opts_count(opts) == 1);
- long_key[127] = 0;
- g_assert_cmpstr(qemu_opt_get(opts, long_key), ==, "v");
- g_free(params);
-
/* Multiple keys, last one wins */
opts = qemu_opts_parse(&opts_list_03, "a=1,b=2,,x,a=3",
false, &error_abort);
diff --git a/util/qemu-option.c b/util/qemu-option.c
index baca40f..fa1a9f1 100644
--- a/util/qemu-option.c
+++ b/util/qemu-option.c
@@ -43,27 +43,23 @@
* first byte of the option name)
*
* The option name is delimited by delim (usually , or =) or the string end
- * and is copied into buf. If the option name is longer than buf_size, it is
- * truncated. buf is always zero terminated.
+ * and is copied into option. The caller is responsible for free'ing option
+ * when no longer required.
*
* The return value is the position of the delimiter/zero byte after the option
* name in p.
*/
-static const char *get_opt_name(char *buf, int buf_size, const char *p,
- char delim)
+static const char *get_opt_name(const char *p, char **option, char delim)
{
- char *q;
+ char *offset = strchr(p, delim);
- q = buf;
- while (*p != '\0' && *p != delim) {
- if (q && (q - buf) < buf_size - 1)
- *q++ = *p;
- p++;
+ if (offset) {
+ *option = g_strndup(p, offset - p);
+ return offset;
+ } else {
+ *option = g_strdup(p);
+ return p + strlen(p);
}
- if (q)
- *q = '\0';
-
- return p;
}
/*
@@ -758,7 +754,8 @@ void qemu_opts_print(QemuOpts *opts, const char *separator)
static void opts_do_parse(QemuOpts *opts, const char *params,
const char *firstname, bool prepend, Error **errp)
{
- char option[128], value[1024];
+ char *option = NULL;
+ char value[1024];
const char *p,*pe,*pc;
Error *local_err = NULL;
@@ -769,11 +766,11 @@ static void opts_do_parse(QemuOpts *opts, const char *params,
/* found "foo,more" */
if (p == params && firstname) {
/* implicitly named first option */
- pstrcpy(option, sizeof(option), firstname);
+ option = g_strdup(firstname);
p = get_opt_value(value, sizeof(value), p);
} else {
/* option without value, probably a flag */
- p = get_opt_name(option, sizeof(option), p, ',');
+ p = get_opt_name(p, &option, ',');
if (strncmp(option, "no", 2) == 0) {
memmove(option, option+2, strlen(option+2)+1);
pstrcpy(value, sizeof(value), "off");
@@ -783,10 +780,8 @@ static void opts_do_parse(QemuOpts *opts, const char *params,
}
} else {
/* found "foo=bar,more" */
- p = get_opt_name(option, sizeof(option), p, '=');
- if (*p != '=') {
- break;
- }
+ p = get_opt_name(p, &option, '=');
+ assert(*p == '=');
p++;
p = get_opt_value(value, sizeof(value), p);
}
@@ -795,13 +790,18 @@ static void opts_do_parse(QemuOpts *opts, const char *params,
opt_set(opts, option, value, prepend, &local_err);
if (local_err) {
error_propagate(errp, local_err);
- return;
+ goto cleanup;
}
}
if (*p != ',') {
break;
}
+ g_free(option);
+ option = NULL;
}
+
+ cleanup:
+ g_free(option);
}
/**
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* Re: [Qemu-devel] [PULL 14/30] opts: don't silently truncate long parameter keys
2018-05-08 22:14 ` [Qemu-devel] [PULL 14/30] opts: don't silently truncate long parameter keys Paolo Bonzini
@ 2018-05-09 5:46 ` Thomas Huth
0 siblings, 0 replies; 40+ messages in thread
From: Thomas Huth @ 2018-05-09 5:46 UTC (permalink / raw)
To: Paolo Bonzini, qemu-devel
On 09.05.2018 00:14, Paolo Bonzini wrote:
> From: Daniel P. Berrangé <berrange@redhat.com>
Here the accent is right ^
> The existing QemuOpts parsing code uses a fixed size 128 byte buffer
> for storing the parameter keys. If a key exceeded this size it was
> silently truncate and no error reported to the user. This behaviour was
> reasonable & harmless because traditionally the key names are all
> statically declared, and it was known that no code was declaring a key
> longer than 127 bytes. This assumption, however, ceased to be valid once
> the block layer added support for dot-separate compound keys. This
> syntax allows for keys that can be arbitrarily long, limited only by the
> number of block drivers you can stack up. With this usage, silently
> truncating the key name can never lead to correct behaviour.
>
> Hopefully such truncation would turn into an error, when the block code
> then tried to extract options later, but there's no guarantee that will
> happen. It is conceivable that an option specified by the user may be
> truncated and then ignored. This could have serious consequences,
> possibly even leading to security problems if the ignored option set a
> security relevant parameter.
>
> If the operating system didn't limit the user's argv when spawning QEMU,
> the code should honour whatever length arguments were given without
> imposing its own length restrictions. This patch thus changes the code
> to use a heap allocated buffer for storing the keys during parsing,
> lifting the arbitrary length restriction.
>
> Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
But this one got mojibaked -------^
Thomas
^ permalink raw reply [flat|nested] 40+ messages in thread
* [Qemu-devel] [PULL 15/30] opts: don't silently truncate long option values
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
` (13 preceding siblings ...)
2018-05-08 22:14 ` [Qemu-devel] [PULL 14/30] opts: don't silently truncate long parameter keys Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-14 16:19 ` Peter Maydell
2018-05-08 22:14 ` [Qemu-devel] [PULL 16/30] target/i386: sev: fix memory leaks Paolo Bonzini
` (15 subsequent siblings)
30 siblings, 1 reply; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel
From: Daniel P. Berrangé <berrange@redhat.com>
The existing QemuOpts parsing code uses a fixed size 1024 byte buffer
for storing the option values. If a value exceeded this size it was
silently truncated and no error reported to the user. Long option values
is not a common scenario, but it is conceivable that they will happen.
eg if the user has a very deeply nested filesystem it would be possible
to come up with a disk path that was > 1024 bytes. Most of the time if
such data was silently truncated, the user would get an error about
opening a non-existant disk. If they're unlucky though, QEMU might use a
completely different disk image from another VM, which could be
considered a security issue. Another example program was in using the
-smbios command line arg with very large data blobs. In this case the
silent truncation will be providing semantically incorrect data to the
guest OS for SMBIOS tables.
If the operating system didn't limit the user's argv when spawning QEMU,
the code should honour whatever length arguments were given without
imposing its own length restrictions. This patch thus changes the code
to use a heap allocated buffer for storing the values during parsing,
lifting the arbitrary length restriction.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Message-Id: <20180416111743.8473-4-berrange@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
---
hw/i386/multiboot.c | 33 +++++++++------
include/qemu/option.h | 2 +-
util/qemu-option.c | 111 +++++++++++++++++++++++++++-----------------------
3 files changed, 81 insertions(+), 65 deletions(-)
diff --git a/hw/i386/multiboot.c b/hw/i386/multiboot.c
index 5bc0a2c..7a2953e 100644
--- a/hw/i386/multiboot.c
+++ b/hw/i386/multiboot.c
@@ -291,12 +291,16 @@ int load_multiboot(FWCfgState *fw_cfg,
cmdline_len = strlen(kernel_filename) + 1;
cmdline_len += strlen(kernel_cmdline) + 1;
if (initrd_filename) {
- const char *r = initrd_filename;
+ const char *r = get_opt_value(initrd_filename, NULL);
cmdline_len += strlen(r) + 1;
mbs.mb_mods_avail = 1;
- while (*(r = get_opt_value(NULL, 0, r))) {
- mbs.mb_mods_avail++;
- r++;
+ while (1) {
+ mbs.mb_mods_avail++;
+ r = get_opt_value(r, NULL);
+ if (!*r) {
+ break;
+ }
+ r++;
}
}
@@ -313,7 +317,8 @@ int load_multiboot(FWCfgState *fw_cfg,
if (initrd_filename) {
const char *next_initrd;
- char not_last, tmpbuf[strlen(initrd_filename) + 1];
+ char not_last;
+ char *one_file = NULL;
mbs.offset_mods = mbs.mb_buf_size;
@@ -322,24 +327,26 @@ int load_multiboot(FWCfgState *fw_cfg,
int mb_mod_length;
uint32_t offs = mbs.mb_buf_size;
- next_initrd = get_opt_value(tmpbuf, sizeof(tmpbuf), initrd_filename);
+ next_initrd = get_opt_value(initrd_filename, &one_file);
not_last = *next_initrd;
/* if a space comes after the module filename, treat everything
after that as parameters */
- hwaddr c = mb_add_cmdline(&mbs, tmpbuf);
- if ((next_space = strchr(tmpbuf, ' ')))
+ hwaddr c = mb_add_cmdline(&mbs, one_file);
+ next_space = strchr(one_file, ' ');
+ if (next_space) {
*next_space = '\0';
- mb_debug("multiboot loading module: %s", tmpbuf);
- mb_mod_length = get_image_size(tmpbuf);
+ }
+ mb_debug("multiboot loading module: %s", one_file);
+ mb_mod_length = get_image_size(one_file);
if (mb_mod_length < 0) {
- error_report("Failed to open file '%s'", tmpbuf);
+ error_report("Failed to open file '%s'", one_file);
exit(1);
}
mbs.mb_buf_size = TARGET_PAGE_ALIGN(mb_mod_length + mbs.mb_buf_size);
mbs.mb_buf = g_realloc(mbs.mb_buf, mbs.mb_buf_size);
- load_image(tmpbuf, (unsigned char *)mbs.mb_buf + offs);
+ load_image(one_file, (unsigned char *)mbs.mb_buf + offs);
mb_add_mod(&mbs, mbs.mb_buf_phys + offs,
mbs.mb_buf_phys + offs + mb_mod_length, c);
@@ -347,6 +354,8 @@ int load_multiboot(FWCfgState *fw_cfg,
(char *)mbs.mb_buf + offs,
(char *)mbs.mb_buf + offs + mb_mod_length, c);
initrd_filename = next_initrd+1;
+ g_free(one_file);
+ one_file = NULL;
} while (not_last);
}
diff --git a/include/qemu/option.h b/include/qemu/option.h
index 1cfe5cb..3dfb449 100644
--- a/include/qemu/option.h
+++ b/include/qemu/option.h
@@ -28,7 +28,7 @@
#include "qemu/queue.h"
-const char *get_opt_value(char *buf, int buf_size, const char *p);
+const char *get_opt_value(const char *p, char **value);
void parse_option_size(const char *name, const char *value,
uint64_t *ret, Error **errp);
diff --git a/util/qemu-option.c b/util/qemu-option.c
index fa1a9f1..58d1c23 100644
--- a/util/qemu-option.c
+++ b/util/qemu-option.c
@@ -70,25 +70,37 @@ static const char *get_opt_name(const char *p, char **option, char delim)
* delimiter is fixed to be comma which starts a new option. To specify an
* option value that contains commas, double each comma.
*/
-const char *get_opt_value(char *buf, int buf_size, const char *p)
+const char *get_opt_value(const char *p, char **value)
{
- char *q;
+ size_t capacity = 0, length;
+ const char *offset;
+
+ *value = NULL;
+ while (1) {
+ offset = strchr(p, ',');
+ if (!offset) {
+ offset = p + strlen(p);
+ }
- q = buf;
- while (*p != '\0') {
- if (*p == ',') {
- if (*(p + 1) != ',')
- break;
- p++;
+ length = offset - p;
+ if (*offset != '\0' && *(offset + 1) == ',') {
+ length++;
+ }
+ if (value) {
+ *value = g_renew(char, *value, capacity + length + 1);
+ strncpy(*value + capacity, p, length);
+ (*value)[capacity + length] = '\0';
+ }
+ capacity += length;
+ if (*offset == '\0' ||
+ *(offset + 1) != ',') {
+ break;
}
- if (q && (q - buf) < buf_size - 1)
- *q++ = *p;
- p++;
+
+ p += (offset - p) + 2;
}
- if (q)
- *q = '\0';
- return p;
+ return offset;
}
static void parse_option_bool(const char *name, const char *value, bool *ret,
@@ -162,50 +174,43 @@ void parse_option_size(const char *name, const char *value,
bool has_help_option(const char *param)
{
- size_t buflen = strlen(param) + 1;
- char *buf = g_malloc(buflen);
const char *p = param;
bool result = false;
- while (*p) {
- p = get_opt_value(buf, buflen, p);
+ while (*p && !result) {
+ char *value;
+
+ p = get_opt_value(p, &value);
if (*p) {
p++;
}
- if (is_help_option(buf)) {
- result = true;
- goto out;
- }
+ result = is_help_option(value);
+ g_free(value);
}
-out:
- g_free(buf);
return result;
}
-bool is_valid_option_list(const char *param)
+bool is_valid_option_list(const char *p)
{
- size_t buflen = strlen(param) + 1;
- char *buf = g_malloc(buflen);
- const char *p = param;
- bool result = true;
+ char *value = NULL;
+ bool result = false;
while (*p) {
- p = get_opt_value(buf, buflen, p);
- if (*p && !*++p) {
- result = false;
+ p = get_opt_value(p, &value);
+ if ((*p && !*++p) ||
+ (!*value || *value == ',')) {
goto out;
}
- if (!*buf || *buf == ',') {
- result = false;
- goto out;
- }
+ g_free(value);
+ value = NULL;
}
+ result = true;
out:
- g_free(buf);
+ g_free(value);
return result;
}
@@ -487,7 +492,7 @@ int qemu_opt_unset(QemuOpts *opts, const char *name)
}
}
-static void opt_set(QemuOpts *opts, const char *name, const char *value,
+static void opt_set(QemuOpts *opts, const char *name, char *value,
bool prepend, Error **errp)
{
QemuOpt *opt;
@@ -496,6 +501,7 @@ static void opt_set(QemuOpts *opts, const char *name, const char *value,
desc = find_desc_by_name(opts->list->desc, name);
if (!desc && !opts_accepts_any(opts)) {
+ g_free(value);
error_setg(errp, QERR_INVALID_PARAMETER, name);
return;
}
@@ -509,8 +515,7 @@ static void opt_set(QemuOpts *opts, const char *name, const char *value,
QTAILQ_INSERT_TAIL(&opts->head, opt, next);
}
opt->desc = desc;
- opt->str = g_strdup(value);
- assert(opt->str);
+ opt->str = value;
qemu_opt_parse(opt, &local_err);
if (local_err) {
error_propagate(errp, local_err);
@@ -521,7 +526,7 @@ static void opt_set(QemuOpts *opts, const char *name, const char *value,
void qemu_opt_set(QemuOpts *opts, const char *name, const char *value,
Error **errp)
{
- opt_set(opts, name, value, false, errp);
+ opt_set(opts, name, g_strdup(value), false, errp);
}
void qemu_opt_set_bool(QemuOpts *opts, const char *name, bool val,
@@ -755,7 +760,7 @@ static void opts_do_parse(QemuOpts *opts, const char *params,
const char *firstname, bool prepend, Error **errp)
{
char *option = NULL;
- char value[1024];
+ char *value = NULL;
const char *p,*pe,*pc;
Error *local_err = NULL;
@@ -767,15 +772,15 @@ static void opts_do_parse(QemuOpts *opts, const char *params,
if (p == params && firstname) {
/* implicitly named first option */
option = g_strdup(firstname);
- p = get_opt_value(value, sizeof(value), p);
+ p = get_opt_value(p, &value);
} else {
/* option without value, probably a flag */
p = get_opt_name(p, &option, ',');
if (strncmp(option, "no", 2) == 0) {
memmove(option, option+2, strlen(option+2)+1);
- pstrcpy(value, sizeof(value), "off");
+ value = g_strdup("off");
} else {
- pstrcpy(value, sizeof(value), "on");
+ value = g_strdup("on");
}
}
} else {
@@ -783,11 +788,12 @@ static void opts_do_parse(QemuOpts *opts, const char *params,
p = get_opt_name(p, &option, '=');
assert(*p == '=');
p++;
- p = get_opt_value(value, sizeof(value), p);
+ p = get_opt_value(p, &value);
}
if (strcmp(option, "id") != 0) {
/* store and parse */
opt_set(opts, option, value, prepend, &local_err);
+ value = NULL;
if (local_err) {
error_propagate(errp, local_err);
goto cleanup;
@@ -797,11 +803,13 @@ static void opts_do_parse(QemuOpts *opts, const char *params,
break;
}
g_free(option);
- option = NULL;
+ g_free(value);
+ option = value = NULL;
}
cleanup:
g_free(option);
+ g_free(value);
}
/**
@@ -820,7 +828,7 @@ static QemuOpts *opts_parse(QemuOptsList *list, const char *params,
bool permit_abbrev, bool defaults, Error **errp)
{
const char *firstname;
- char value[1024], *id = NULL;
+ char *id = NULL;
const char *p;
QemuOpts *opts;
Error *local_err = NULL;
@@ -829,11 +837,9 @@ static QemuOpts *opts_parse(QemuOptsList *list, const char *params,
firstname = permit_abbrev ? list->implied_opt_name : NULL;
if (strncmp(params, "id=", 3) == 0) {
- get_opt_value(value, sizeof(value), params+3);
- id = value;
+ get_opt_value(params + 3, &id);
} else if ((p = strstr(params, ",id=")) != NULL) {
- get_opt_value(value, sizeof(value), p+4);
- id = value;
+ get_opt_value(p + 4, &id);
}
/*
@@ -845,6 +851,7 @@ static QemuOpts *opts_parse(QemuOptsList *list, const char *params,
*/
assert(!defaults || list->merge_lists);
opts = qemu_opts_create(list, id, !defaults, &local_err);
+ g_free(id);
if (opts == NULL) {
error_propagate(errp, local_err);
return NULL;
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* Re: [Qemu-devel] [PULL 15/30] opts: don't silently truncate long option values
2018-05-08 22:14 ` [Qemu-devel] [PULL 15/30] opts: don't silently truncate long option values Paolo Bonzini
@ 2018-05-14 16:19 ` Peter Maydell
2018-05-14 16:23 ` Daniel P. Berrangé
0 siblings, 1 reply; 40+ messages in thread
From: Peter Maydell @ 2018-05-14 16:19 UTC (permalink / raw)
To: Paolo Bonzini; +Cc: QEMU Developers
On 8 May 2018 at 23:14, Paolo Bonzini <pbonzini@redhat.com> wrote:
> From: Daniel P. Berrangé <berrange@redhat.com>
>
> The existing QemuOpts parsing code uses a fixed size 1024 byte buffer
> for storing the option values. If a value exceeded this size it was
> silently truncated and no error reported to the user. Long option values
> is not a common scenario, but it is conceivable that they will happen.
> eg if the user has a very deeply nested filesystem it would be possible
> to come up with a disk path that was > 1024 bytes. Most of the time if
> such data was silently truncated, the user would get an error about
> opening a non-existant disk. If they're unlucky though, QEMU might use a
> completely different disk image from another VM, which could be
> considered a security issue. Another example program was in using the
> -smbios command line arg with very large data blobs. In this case the
> silent truncation will be providing semantically incorrect data to the
> guest OS for SMBIOS tables.
>
> If the operating system didn't limit the user's argv when spawning QEMU,
> the code should honour whatever length arguments were given without
> imposing its own length restrictions. This patch thus changes the code
> to use a heap allocated buffer for storing the values during parsing,
> lifting the arbitrary length restriction.
Hi; Coverity doesn't like this change (CID1391003):
> --- a/util/qemu-option.c
> +++ b/util/qemu-option.c
> @@ -70,25 +70,37 @@ static const char *get_opt_name(const char *p, char **option, char delim)
> * delimiter is fixed to be comma which starts a new option. To specify an
> * option value that contains commas, double each comma.
> */
> -const char *get_opt_value(char *buf, int buf_size, const char *p)
> +const char *get_opt_value(const char *p, char **value)
> {
> - char *q;
> + size_t capacity = 0, length;
> + const char *offset;
> +
> + *value = NULL;
Here we write to *value, so value must be non-NULL, and
within the loop the only place we write to value it
can't become NULL either (g_renew can't fail)...
> + while (1) {
> + offset = strchr(p, ',');
> + if (!offset) {
> + offset = p + strlen(p);
> + }
>
> - q = buf;
> - while (*p != '\0') {
> - if (*p == ',') {
> - if (*(p + 1) != ',')
> - break;
> - p++;
> + length = offset - p;
> + if (*offset != '\0' && *(offset + 1) == ',') {
> + length++;
> + }
> + if (value) {
...so this check for whether value is NULL can never be true.
> + *value = g_renew(char, *value, capacity + length + 1);
> + strncpy(*value + capacity, p, length);
> + (*value)[capacity + length] = '\0';
> + }
> + capacity += length;
> + if (*offset == '\0' ||
> + *(offset + 1) != ',') {
> + break;
> }
> - if (q && (q - buf) < buf_size - 1)
> - *q++ = *p;
> - p++;
> +
> + p += (offset - p) + 2;
> }
> - if (q)
> - *q = '\0';
>
> - return p;
> + return offset;
> }
>
thanks
-- PMM
^ permalink raw reply [flat|nested] 40+ messages in thread* Re: [Qemu-devel] [PULL 15/30] opts: don't silently truncate long option values
2018-05-14 16:19 ` Peter Maydell
@ 2018-05-14 16:23 ` Daniel P. Berrangé
0 siblings, 0 replies; 40+ messages in thread
From: Daniel P. Berrangé @ 2018-05-14 16:23 UTC (permalink / raw)
To: Peter Maydell; +Cc: Paolo Bonzini, QEMU Developers
On Mon, May 14, 2018 at 05:19:04PM +0100, Peter Maydell wrote:
> On 8 May 2018 at 23:14, Paolo Bonzini <pbonzini@redhat.com> wrote:
> > From: Daniel P. Berrangé <berrange@redhat.com>
> >
> > The existing QemuOpts parsing code uses a fixed size 1024 byte buffer
> > for storing the option values. If a value exceeded this size it was
> > silently truncated and no error reported to the user. Long option values
> > is not a common scenario, but it is conceivable that they will happen.
> > eg if the user has a very deeply nested filesystem it would be possible
> > to come up with a disk path that was > 1024 bytes. Most of the time if
> > such data was silently truncated, the user would get an error about
> > opening a non-existant disk. If they're unlucky though, QEMU might use a
> > completely different disk image from another VM, which could be
> > considered a security issue. Another example program was in using the
> > -smbios command line arg with very large data blobs. In this case the
> > silent truncation will be providing semantically incorrect data to the
> > guest OS for SMBIOS tables.
> >
> > If the operating system didn't limit the user's argv when spawning QEMU,
> > the code should honour whatever length arguments were given without
> > imposing its own length restrictions. This patch thus changes the code
> > to use a heap allocated buffer for storing the values during parsing,
> > lifting the arbitrary length restriction.
>
> Hi; Coverity doesn't like this change (CID1391003):
>
> > --- a/util/qemu-option.c
> > +++ b/util/qemu-option.c
> > @@ -70,25 +70,37 @@ static const char *get_opt_name(const char *p, char **option, char delim)
> > * delimiter is fixed to be comma which starts a new option. To specify an
> > * option value that contains commas, double each comma.
> > */
> > -const char *get_opt_value(char *buf, int buf_size, const char *p)
> > +const char *get_opt_value(const char *p, char **value)
> > {
> > - char *q;
> > + size_t capacity = 0, length;
> > + const char *offset;
> > +
> > + *value = NULL;
>
> Here we write to *value, so value must be non-NULL, and
> within the loop the only place we write to value it
> can't become NULL either (g_renew can't fail)...
Oh, real bug ! This should have been
if (value) {
*value = NULL;
}
because multiboot.c passes in NULL for this parameter.
Unless we decide to rewrite multiboot.c to avoid that instead,
since all other callers pass non-NULL.
>
> > + while (1) {
> > + offset = strchr(p, ',');
> > + if (!offset) {
> > + offset = p + strlen(p);
> > + }
> >
> > - q = buf;
> > - while (*p != '\0') {
> > - if (*p == ',') {
> > - if (*(p + 1) != ',')
> > - break;
> > - p++;
> > + length = offset - p;
> > + if (*offset != '\0' && *(offset + 1) == ',') {
> > + length++;
> > + }
> > + if (value) {
>
> ...so this check for whether value is NULL can never be true.
>
> > + *value = g_renew(char, *value, capacity + length + 1);
> > + strncpy(*value + capacity, p, length);
> > + (*value)[capacity + length] = '\0';
> > + }
> > + capacity += length;
> > + if (*offset == '\0' ||
> > + *(offset + 1) != ',') {
> > + break;
> > }
> > - if (q && (q - buf) < buf_size - 1)
> > - *q++ = *p;
> > - p++;
> > +
> > + p += (offset - p) + 2;
> > }
> > - if (q)
> > - *q = '\0';
> >
> > - return p;
> > + return offset;
> > }
> >
>
> thanks
> -- PMM
>
Regards,
Daniel
--
|: https://berrange.com -o- https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org -o- https://fstop138.berrange.com :|
|: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|
^ permalink raw reply [flat|nested] 40+ messages in thread
* [Qemu-devel] [PULL 16/30] target/i386: sev: fix memory leaks
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
` (14 preceding siblings ...)
2018-05-08 22:14 ` [Qemu-devel] [PULL 15/30] opts: don't silently truncate long option values Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 17/30] qemu-options: Mark -virtioconsole as deprecated Paolo Bonzini
` (14 subsequent siblings)
30 siblings, 0 replies; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel
Reported by Coverity.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
target/i386/sev.c | 32 +++++++++++++++++---------------
1 file changed, 17 insertions(+), 15 deletions(-)
diff --git a/target/i386/sev.c b/target/i386/sev.c
index c011671..2395171 100644
--- a/target/i386/sev.c
+++ b/target/i386/sev.c
@@ -430,7 +430,8 @@ static int
sev_get_pdh_info(int fd, guchar **pdh, size_t *pdh_len, guchar **cert_chain,
size_t *cert_chain_len)
{
- guchar *pdh_data, *cert_chain_data;
+ guchar *pdh_data = NULL;
+ guchar *cert_chain_data = NULL;
struct sev_user_data_pdh_cert_export export = {};
int err, r;
@@ -471,8 +472,9 @@ e_free:
SevCapability *
sev_get_capabilities(void)
{
- SevCapability *cap;
- guchar *pdh_data, *cert_chain_data;
+ SevCapability *cap = NULL;
+ guchar *pdh_data = NULL;
+ guchar *cert_chain_data = NULL;
size_t pdh_len = 0, cert_chain_len = 0;
uint32_t ebx;
int fd;
@@ -486,7 +488,7 @@ sev_get_capabilities(void)
if (sev_get_pdh_info(fd, &pdh_data, &pdh_len,
&cert_chain_data, &cert_chain_len)) {
- return NULL;
+ goto out;
}
cap = g_new0(SevCapability, 1);
@@ -502,9 +504,9 @@ sev_get_capabilities(void)
*/
cap->reduced_phys_bits = 1;
+out:
g_free(pdh_data);
g_free(cert_chain_data);
-
close(fd);
return cap;
}
@@ -530,7 +532,7 @@ sev_launch_start(SEVState *s)
{
gsize sz;
int ret = 1;
- int fw_error;
+ int fw_error, rc;
QSevGuestInfo *sev = s->sev_info;
struct kvm_sev_launch_start *start;
guchar *session = NULL, *dh_cert = NULL;
@@ -543,7 +545,7 @@ sev_launch_start(SEVState *s)
&error_abort);
if (sev->session_file) {
if (sev_read_file_base64(sev->session_file, &session, &sz) < 0) {
- return 1;
+ goto out;
}
start->session_uaddr = (unsigned long)session;
start->session_len = sz;
@@ -551,18 +553,18 @@ sev_launch_start(SEVState *s)
if (sev->dh_cert_file) {
if (sev_read_file_base64(sev->dh_cert_file, &dh_cert, &sz) < 0) {
- return 1;
+ goto out;
}
start->dh_uaddr = (unsigned long)dh_cert;
start->dh_len = sz;
}
trace_kvm_sev_launch_start(start->policy, session, dh_cert);
- ret = sev_ioctl(s->sev_fd, KVM_SEV_LAUNCH_START, start, &fw_error);
- if (ret < 0) {
+ rc = sev_ioctl(s->sev_fd, KVM_SEV_LAUNCH_START, start, &fw_error);
+ if (rc < 0) {
error_report("%s: LAUNCH_START ret=%d fw_error=%d '%s'",
__func__, ret, fw_error, fw_error_to_str(fw_error));
- return 1;
+ goto out;
}
object_property_set_int(OBJECT(sev), start->handle, "handle",
@@ -570,12 +572,13 @@ sev_launch_start(SEVState *s)
sev_set_guest_state(SEV_STATE_LAUNCH_UPDATE);
s->handle = start->handle;
s->policy = start->policy;
+ ret = 0;
+out:
g_free(start);
g_free(session);
g_free(dh_cert);
-
- return 0;
+ return ret;
}
static int
@@ -712,7 +715,7 @@ sev_guest_init(const char *id)
uint32_t host_cbitpos;
struct sev_user_data_status status = {};
- s = g_new0(SEVState, 1);
+ sev_state = s = g_new0(SEVState, 1);
s->sev_info = lookup_sev_guest_info(id);
if (!s->sev_info) {
error_report("%s: '%s' is not a valid '%s' object",
@@ -720,7 +723,6 @@ sev_guest_init(const char *id)
goto err;
}
- sev_state = s;
s->state = SEV_STATE_UNINIT;
host_cpuid(0x8000001F, 0, NULL, &ebx, NULL, NULL);
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* [Qemu-devel] [PULL 17/30] qemu-options: Mark -virtioconsole as deprecated
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
` (15 preceding siblings ...)
2018-05-08 22:14 ` [Qemu-devel] [PULL 16/30] target/i386: sev: fix memory leaks Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 18/30] qemu-options: Remove remainders of the -tdf option Paolo Bonzini
` (13 subsequent siblings)
30 siblings, 0 replies; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel; +Cc: Thomas Huth
From: Thomas Huth <thuth@redhat.com>
The qemu-doc already states that this option is only maintained for
backward compatibility and "-device virtconsole" should be used
instead. So let's take the next step and mark this option officially
as deprecated.
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-Id: <1525446790-16139-1-git-send-email-thuth@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
qemu-doc.texi | 5 +++++
qemu-options.hx | 5 +----
vl.c | 2 ++
3 files changed, 8 insertions(+), 4 deletions(-)
diff --git a/qemu-doc.texi b/qemu-doc.texi
index 0ed0f19..4add403 100644
--- a/qemu-doc.texi
+++ b/qemu-doc.texi
@@ -2931,6 +2931,11 @@ The @code{-localtime} option has been replaced by @code{-rtc base=localtime}.
The @code{-startdate} option has been replaced by @code{-rtc base=@var{date}}.
+@subsection -virtioconsole (since 2.13.0)
+
+Option @option{-virtioconsole} has been replaced by
+@option{-device virtconsole}.
+
@section qemu-img command line arguments
@subsection convert -s (since 2.0.0)
diff --git a/qemu-options.hx b/qemu-options.hx
index c611766..091ded6 100644
--- a/qemu-options.hx
+++ b/qemu-options.hx
@@ -3675,10 +3675,7 @@ STEXI
@item -virtioconsole @var{c}
@findex -virtioconsole
Set virtio console.
-
-This option is maintained for backward compatibility.
-
-Please use @code{-device virtconsole} for the new way of invocation.
+This option is deprecated, please use @option{-device virtconsole} instead.
ETEXI
DEF("show-cursor", 0, QEMU_OPTION_show_cursor, \
diff --git a/vl.c b/vl.c
index 12e31d1..b928e3e 100644
--- a/vl.c
+++ b/vl.c
@@ -3587,6 +3587,8 @@ int main(int argc, char **argv, char **envp)
}
break;
case QEMU_OPTION_virtiocon:
+ warn_report("This option is deprecated, "
+ "use '-device virtconsole' instead");
add_device_config(DEV_VIRTCON, optarg);
default_virtcon = 0;
if (strncmp(optarg, "mon:", 4) == 0) {
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* [Qemu-devel] [PULL 18/30] qemu-options: Remove remainders of the -tdf option
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
` (16 preceding siblings ...)
2018-05-08 22:14 ` [Qemu-devel] [PULL 17/30] qemu-options: Mark -virtioconsole as deprecated Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 19/30] qemu-options: Bail out on unsupported options instead of silently ignoring them Paolo Bonzini
` (12 subsequent siblings)
30 siblings, 0 replies; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel; +Cc: Thomas Huth
From: Thomas Huth <thuth@redhat.com>
The -tdf options has been removed with d07aa197c5a1556449361a0cbb5108e2,
but apparently I forgot to remove the corresponding two lines from
qemu-options.hx, so this option is still "available" and just silently
ignored. Kill it now for good.
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-Id: <1525453270-23074-2-git-send-email-thuth@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
qemu-options.hx | 3 ---
1 file changed, 3 deletions(-)
diff --git a/qemu-options.hx b/qemu-options.hx
index 091ded6..50c906c 100644
--- a/qemu-options.hx
+++ b/qemu-options.hx
@@ -3921,9 +3921,6 @@ DEF("no-kvm-pit-reinjection", 0, QEMU_OPTION_no_kvm_pit_reinjection,
HXCOMM Deprecated by -machine kernel_irqchip=on|off property
DEF("no-kvm-irqchip", 0, QEMU_OPTION_no_kvm_irqchip, "", QEMU_ARCH_I386)
-HXCOMM Deprecated (ignored)
-DEF("tdf", 0, QEMU_OPTION_tdf,"", QEMU_ARCH_ALL)
-
DEF("msg", HAS_ARG, QEMU_OPTION_msg,
"-msg timestamp[=on|off]\n"
" change the format of messages\n"
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* [Qemu-devel] [PULL 19/30] qemu-options: Bail out on unsupported options instead of silently ignoring them
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
` (17 preceding siblings ...)
2018-05-08 22:14 ` [Qemu-devel] [PULL 18/30] qemu-options: Remove remainders of the -tdf option Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 20/30] qemu-options: Remove deprecated -no-kvm-pit-reinjection Paolo Bonzini
` (11 subsequent siblings)
30 siblings, 0 replies; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel; +Cc: Thomas Huth
From: Thomas Huth <thuth@redhat.com>
The dangling remainder of the -tdf option revealed a deficiency in our
option parsing: Options that have been declared, but are not supported
in the switch-case statement in vl.c and not handled in the OS-specifc
os_parse_cmd_args() functions are currently silently ignored. We should
rather tell the users that they specified something that we can not
handle, so let's print an error message and exit instead.
Reported-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-Id: <1525453270-23074-3-git-send-email-thuth@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
include/qemu-common.h | 2 +-
os-posix.c | 6 +++++-
os-win32.c | 4 ++--
vl.c | 5 ++++-
4 files changed, 12 insertions(+), 5 deletions(-)
diff --git a/include/qemu-common.h b/include/qemu-common.h
index 8a4f63c..85f4749 100644
--- a/include/qemu-common.h
+++ b/include/qemu-common.h
@@ -137,7 +137,7 @@ char *qemu_find_file(int type, const char *name);
/* OS specific functions */
void os_setup_early_signal_handling(void);
char *os_find_datadir(void);
-void os_parse_cmd_args(int index, const char *optarg);
+int os_parse_cmd_args(int index, const char *optarg);
#include "qemu/module.h"
diff --git a/os-posix.c b/os-posix.c
index 24eb700..9ce6f74 100644
--- a/os-posix.c
+++ b/os-posix.c
@@ -165,7 +165,7 @@ static bool os_parse_runas_uid_gid(const char *optarg)
* Parse OS specific command line options.
* return 0 if option handled, -1 otherwise
*/
-void os_parse_cmd_args(int index, const char *optarg)
+int os_parse_cmd_args(int index, const char *optarg)
{
switch (index) {
#ifdef CONFIG_SLIRP
@@ -199,7 +199,11 @@ void os_parse_cmd_args(int index, const char *optarg)
fips_set_state(true);
break;
#endif
+ default:
+ return -1;
}
+
+ return 0;
}
static void change_process_uid(void)
diff --git a/os-win32.c b/os-win32.c
index 586a7c7..0674f94 100644
--- a/os-win32.c
+++ b/os-win32.c
@@ -93,9 +93,9 @@ void os_set_line_buffering(void)
* Parse OS specific command line options.
* return 0 if option handled, -1 otherwise
*/
-void os_parse_cmd_args(int index, const char *optarg)
+int os_parse_cmd_args(int index, const char *optarg)
{
- return;
+ return -1;
}
int qemu_create_pidfile(const char *filename)
diff --git a/vl.c b/vl.c
index b928e3e..a23acb2 100644
--- a/vl.c
+++ b/vl.c
@@ -4033,7 +4033,10 @@ int main(int argc, char **argv, char **envp)
}
break;
default:
- os_parse_cmd_args(popt->index, optarg);
+ if (os_parse_cmd_args(popt->index, optarg)) {
+ error_report("Option not supported in this build");
+ exit(1);
+ }
}
}
}
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* [Qemu-devel] [PULL 20/30] qemu-options: Remove deprecated -no-kvm-pit-reinjection
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
` (18 preceding siblings ...)
2018-05-08 22:14 ` [Qemu-devel] [PULL 19/30] qemu-options: Bail out on unsupported options instead of silently ignoring them Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 21/30] qemu-options: Remove deprecated -no-kvm-irqchip Paolo Bonzini
` (10 subsequent siblings)
30 siblings, 0 replies; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel; +Cc: Thomas Huth
From: Thomas Huth <thuth@redhat.com>
Deprecated since the beginning when it was added for compatibility with
the ancient qemu-kvm fork of QEMU, and it even printed out the deprecation
warning since right from the start (i.e. QEMU v1.3.0), so it's really time
to remove this now.
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-Id: <1525453270-23074-4-git-send-email-thuth@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
qemu-doc.texi | 5 -----
qemu-options.hx | 4 ----
vl.c | 12 ------------
3 files changed, 21 deletions(-)
diff --git a/qemu-doc.texi b/qemu-doc.texi
index 4add403..85a7a39 100644
--- a/qemu-doc.texi
+++ b/qemu-doc.texi
@@ -2786,11 +2786,6 @@ which is the default.
@section System emulator command line arguments
-@subsection -no-kvm-pit-reinjection (since 1.3.0)
-
-The ``-no-kvm-pit-reinjection'' argument is now a
-synonym for setting ``-global kvm-pit.lost_tick_policy=discard''.
-
@subsection -no-kvm-irqchip (since 1.3.0)
The ``-no-kvm-irqchip'' argument is now a synonym for
diff --git a/qemu-options.hx b/qemu-options.hx
index 50c906c..dfff698 100644
--- a/qemu-options.hx
+++ b/qemu-options.hx
@@ -3914,10 +3914,6 @@ ETEXI
HXCOMM Deprecated by -machine accel=tcg property
DEF("no-kvm", 0, QEMU_OPTION_no_kvm, "", QEMU_ARCH_I386)
-HXCOMM Deprecated by kvm-pit driver properties
-DEF("no-kvm-pit-reinjection", 0, QEMU_OPTION_no_kvm_pit_reinjection,
- "", QEMU_ARCH_I386)
-
HXCOMM Deprecated by -machine kernel_irqchip=on|off property
DEF("no-kvm-irqchip", 0, QEMU_OPTION_no_kvm_irqchip, "", QEMU_ARCH_I386)
diff --git a/vl.c b/vl.c
index a23acb2..23d55a4 100644
--- a/vl.c
+++ b/vl.c
@@ -3696,18 +3696,6 @@ int main(int argc, char **argv, char **envp)
olist = qemu_find_opts("machine");
qemu_opts_parse_noisily(olist, "accel=tcg", false);
break;
- case QEMU_OPTION_no_kvm_pit_reinjection: {
- static GlobalProperty kvm_pit_lost_tick_policy = {
- .driver = "kvm-pit",
- .property = "lost_tick_policy",
- .value = "discard",
- };
-
- warn_report("deprecated, replaced by "
- "-global kvm-pit.lost_tick_policy=discard");
- qdev_prop_register_global(&kvm_pit_lost_tick_policy);
- break;
- }
case QEMU_OPTION_accel:
accel_opts = qemu_opts_parse_noisily(qemu_find_opts("accel"),
optarg, true);
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* [Qemu-devel] [PULL 21/30] qemu-options: Remove deprecated -no-kvm-irqchip
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
` (19 preceding siblings ...)
2018-05-08 22:14 ` [Qemu-devel] [PULL 20/30] qemu-options: Remove deprecated -no-kvm-pit-reinjection Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 22/30] qemu-doc: provide details of supported build platforms Paolo Bonzini
` (9 subsequent siblings)
30 siblings, 0 replies; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel; +Cc: Thomas Huth
From: Thomas Huth <thuth@redhat.com>
We've never documented this option in our qemu-doc, so apart from the users
that already used the old qemu-kvm fork before, most users should not be
aware of this option at all. It's been marked as deprecated in the source
code for a long time already, and officially marked as deprecated in the
documentation since QEMU v2.10, so it should be fine to remove this now.
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-Id: <1525453270-23074-5-git-send-email-thuth@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
qemu-doc.texi | 5 -----
qemu-options.hx | 3 ---
vl.c | 5 -----
3 files changed, 13 deletions(-)
diff --git a/qemu-doc.texi b/qemu-doc.texi
index 85a7a39..bde542b 100644
--- a/qemu-doc.texi
+++ b/qemu-doc.texi
@@ -2786,11 +2786,6 @@ which is the default.
@section System emulator command line arguments
-@subsection -no-kvm-irqchip (since 1.3.0)
-
-The ``-no-kvm-irqchip'' argument is now a synonym for
-setting ``-machine kernel_irqchip=off''.
-
@subsection -no-kvm (since 1.3.0)
The ``-no-kvm'' argument is now a synonym for setting
diff --git a/qemu-options.hx b/qemu-options.hx
index dfff698..b2fefd1 100644
--- a/qemu-options.hx
+++ b/qemu-options.hx
@@ -3914,9 +3914,6 @@ ETEXI
HXCOMM Deprecated by -machine accel=tcg property
DEF("no-kvm", 0, QEMU_OPTION_no_kvm, "", QEMU_ARCH_I386)
-HXCOMM Deprecated by -machine kernel_irqchip=on|off property
-DEF("no-kvm-irqchip", 0, QEMU_OPTION_no_kvm_irqchip, "", QEMU_ARCH_I386)
-
DEF("msg", HAS_ARG, QEMU_OPTION_msg,
"-msg timestamp[=on|off]\n"
" change the format of messages\n"
diff --git a/vl.c b/vl.c
index 23d55a4..b9f6b42 100644
--- a/vl.c
+++ b/vl.c
@@ -3138,11 +3138,6 @@ int main(int argc, char **argv, char **envp)
exit(1);
}
switch(popt->index) {
- case QEMU_OPTION_no_kvm_irqchip: {
- olist = qemu_find_opts("machine");
- qemu_opts_parse_noisily(olist, "kernel_irqchip=off", false);
- break;
- }
case QEMU_OPTION_cpu:
/* hw initialization will check this */
cpu_model = optarg;
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* [Qemu-devel] [PULL 22/30] qemu-doc: provide details of supported build platforms
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
` (20 preceding siblings ...)
2018-05-08 22:14 ` [Qemu-devel] [PULL 21/30] qemu-options: Remove deprecated -no-kvm-irqchip Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 23/30] glib: bump min required glib library version to 2.42 Paolo Bonzini
` (8 subsequent siblings)
30 siblings, 0 replies; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel
From: Daniel P. Berrangé <berrange@redhat.com>
Describe the policy the project uses to decide which OS are supported as
build platforms. This will:
- Allow maintainers to determine when the minimum version of a 3rd
party piece of software can be increased without negatively
impacting supported platforms.
- Allow tailoring of CI environments to match the intended supported
build platforms.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Message-Id: <20180504160026.14017-2-berrange@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
---
qemu-doc.texi | 68 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 68 insertions(+)
diff --git a/qemu-doc.texi b/qemu-doc.texi
index bde542b..715bd33 100644
--- a/qemu-doc.texi
+++ b/qemu-doc.texi
@@ -39,6 +39,7 @@
* QEMU User space emulator::
* Implementation notes::
* Deprecated features::
+* Supported build platforms::
* License::
* Index::
@end menu
@@ -2976,6 +2977,73 @@ The ``xlnx-zcu102'' machine has the same features and capabilites in QEMU.
In order to prevent QEMU from automatically opening an image's backing
chain, use ``"backing": null'' instead.
+@node Supported build platforms
+@appendix Supported build platforms
+
+QEMU aims to support building and executing on multiple host OS platforms.
+This appendix outlines which platforms are the major build targets. These
+platforms are used as the basis for deciding upon the minimum required
+versions of 3rd party software QEMU depends on. The supported platforms
+are the targets for automated testing performed by the project when patches
+are submitted for review, and tested before and after merge.
+
+If a platform is not listed here, it does not imply that QEMU won't work.
+If an unlisted platform has comparable software versions to a listed platform,
+there is every expectation that it will work. Bug reports are welcome for
+problems encountered on unlisted platforms unless they are clearly older
+vintage than what is described here.
+
+Note that when considering software versions shipped in distros as support
+targets, QEMU considers only the version number, and assumes the features in
+that distro match the upstream release with the same version. In other words,
+if a distro backports extra features to the software in their distro, QEMU
+upstream code will not add explicit support for those backports, unless the
+feature is auto-detectable in a manner that works for the upstream releases
+too.
+
+The Repology site @url{https://repology.org} is a useful resource to identify
+currently shipped versions of software in various operating systems, though
+it does not cover all distros listed below.
+
+@section Linux OS
+
+For distributions with frequent, short-lifetime releases, the project will
+aim to support all versions that are not end of life by their respective
+vendors. For the purposes of identifying supported software versions, the
+project will look at Fedora, Ubuntu, and openSUSE distros. Other short-
+lifetime distros will be assumed to ship similar software versions.
+
+For distributions with long-lifetime releases, the project will aim to support
+the most recent major version at all times. Support for the previous major
+version will be dropped 2 years after the new major version is released. For
+the purposes of identifying supported software versions, the project will look
+at RHEL, Debian, Ubuntu LTS, and SLES distros. Other long-lifetime distros will
+be assumed to ship similar software versions.
+
+@section Windows
+
+The project supports building with current versions of the MinGW toolchain,
+hosted on Linux.
+
+@section macOS
+
+The project supports building with the two most recent versions of macOS, with
+the current homebrew package set available.
+
+@section FreeBSD
+
+The project aims to support the all the versions which are not end of life.
+
+@section NetBSD
+
+The project aims to support the most recent major version at all times. Support
+for the previous major version will be dropped 2 years after the new major
+version is released.
+
+@section OpenBSD
+
+The project aims to support the all the versions which are not end of life.
+
@node License
@appendix License
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* [Qemu-devel] [PULL 23/30] glib: bump min required glib library version to 2.42
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
` (21 preceding siblings ...)
2018-05-08 22:14 ` [Qemu-devel] [PULL 22/30] qemu-doc: provide details of supported build platforms Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 24/30] i386/kvm: add support for Hyper-V reenlightenment MSRs Paolo Bonzini
` (7 subsequent siblings)
30 siblings, 0 replies; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel
From: Daniel P. Berrangé <berrange@redhat.com>
Per supported platforms doc, the various min glib on relevant distros is:
RHEL-7: 2.50.3
Debian (Stretch): 2.50.3
Debian (Jessie): 2.42.1
OpenBSD (Ports): 2.54.3
FreeBSD (Ports): 2.50.3
OpenSUSE Leap 15: 2.54.3
Ubuntu (Xenial): 2.48.0
macOS (Homebrew): 2.56.0
This suggests that a minimum glib of 2.42 is a reasonable target
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Message-Id: <20180504160026.14017-3-berrange@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
---
configure | 6 +-
docs/devel/testing.rst | 4 +-
include/glib-compat.h | 319 -------------------------------
tests/docker/dockerfiles/centos6.docker | 30 ---
tests/docker/dockerfiles/min-glib.docker | 8 -
tests/test-qmp-event.c | 2 +-
tests/tpm-emu.h | 4 +-
tests/vhost-user-test.c | 4 +-
trace/simple.c | 6 +-
9 files changed, 11 insertions(+), 372 deletions(-)
delete mode 100644 tests/docker/dockerfiles/centos6.docker
delete mode 100644 tests/docker/dockerfiles/min-glib.docker
diff --git a/configure b/configure
index 83a6080..3ff4d77 100755
--- a/configure
+++ b/configure
@@ -3394,11 +3394,7 @@ fi
##########################################
# glib support probe
-if test "$mingw32" = yes; then
- glib_req_ver=2.30
-else
- glib_req_ver=2.22
-fi
+glib_req_ver=2.42
glib_modules=gthread-2.0
if test "$modules" = yes; then
glib_modules="$glib_modules gmodule-export-2.0"
diff --git a/docs/devel/testing.rst b/docs/devel/testing.rst
index 0ca1a2d..f7b7552 100644
--- a/docs/devel/testing.rst
+++ b/docs/devel/testing.rst
@@ -297,9 +297,9 @@ the container, with parameters defined by the make target):
.. code::
- make docker-test-build@min-glib
+ make docker-test-build@centos7
-This will create a container instance using the ``min-glib`` image (the image
+This will create a container instance using the ``centos7`` image (the image
is downloaded and initialized automatically), in which the ``test-build`` job
is executed.
diff --git a/include/glib-compat.h b/include/glib-compat.h
index c49cf87..3b340ab 100644
--- a/include/glib-compat.h
+++ b/include/glib-compat.h
@@ -18,27 +18,6 @@
#include <glib.h>
-/* GLIB version compatibility flags */
-#if !GLIB_CHECK_VERSION(2, 26, 0)
-#define G_TIME_SPAN_SECOND (G_GINT64_CONSTANT(1000000))
-#endif
-
-#if !GLIB_CHECK_VERSION(2, 28, 0)
-static inline gint64 qemu_g_get_monotonic_time(void)
-{
- /* g_get_monotonic_time() is best-effort so we can use the wall clock as a
- * fallback.
- */
-
- GTimeVal time;
- g_get_current_time(&time);
-
- return time.tv_sec * G_TIME_SPAN_SECOND + time.tv_usec;
-}
-/* work around distro backports of this interface */
-#define g_get_monotonic_time() qemu_g_get_monotonic_time()
-#endif
-
#if defined(_WIN32) && !GLIB_CHECK_VERSION(2, 50, 0)
/*
* g_poll has a problem on Windows when using
@@ -48,228 +27,6 @@ static inline gint64 qemu_g_get_monotonic_time(void)
gint g_poll_fixed(GPollFD *fds, guint nfds, gint timeout);
#endif
-#if !GLIB_CHECK_VERSION(2, 30, 0)
-/* Not a 100% compatible implementation, but good enough for most
- * cases. Placeholders are only supported at the end of the
- * template. */
-static inline gchar *qemu_g_dir_make_tmp(gchar const *tmpl, GError **error)
-{
- gchar *path = g_build_filename(g_get_tmp_dir(), tmpl ?: ".XXXXXX", NULL);
-
- if (mkdtemp(path) != NULL) {
- return path;
- }
- /* Error occurred, clean up. */
- g_set_error(error, G_FILE_ERROR, g_file_error_from_errno(errno),
- "mkdtemp() failed");
- g_free(path);
- return NULL;
-}
-#define g_dir_make_tmp(tmpl, error) qemu_g_dir_make_tmp(tmpl, error)
-#endif /* glib 2.30 */
-
-#if !GLIB_CHECK_VERSION(2, 31, 0)
-/* before glib-2.31, GMutex and GCond was dynamic-only (there was a separate
- * GStaticMutex, but it didn't work with condition variables).
- *
- * Our implementation uses GOnce to fake a static implementation that does
- * not require separate initialization.
- * We need to rename the types to avoid passing our CompatGMutex/CompatGCond
- * by mistake to a function that expects GMutex/GCond. However, for ease
- * of use we keep the GLib function names. GLib uses macros for the
- * implementation, we use inline functions instead and undefine the macros.
- */
-
-typedef struct CompatGMutex {
- GOnce once;
-} CompatGMutex;
-
-typedef struct CompatGCond {
- GOnce once;
-} CompatGCond;
-
-static inline gpointer do_g_mutex_new(gpointer unused)
-{
- return (gpointer) g_mutex_new();
-}
-
-static inline void g_mutex_init(CompatGMutex *mutex)
-{
- mutex->once = (GOnce) G_ONCE_INIT;
-}
-
-static inline void g_mutex_clear(CompatGMutex *mutex)
-{
- g_assert(mutex->once.status != G_ONCE_STATUS_PROGRESS);
- if (mutex->once.retval) {
- g_mutex_free((GMutex *) mutex->once.retval);
- }
- mutex->once = (GOnce) G_ONCE_INIT;
-}
-
-static inline void (g_mutex_lock)(CompatGMutex *mutex)
-{
- g_once(&mutex->once, do_g_mutex_new, NULL);
- g_mutex_lock((GMutex *) mutex->once.retval);
-}
-#undef g_mutex_lock
-
-static inline gboolean (g_mutex_trylock)(CompatGMutex *mutex)
-{
- g_once(&mutex->once, do_g_mutex_new, NULL);
- return g_mutex_trylock((GMutex *) mutex->once.retval);
-}
-#undef g_mutex_trylock
-
-
-static inline void (g_mutex_unlock)(CompatGMutex *mutex)
-{
- g_mutex_unlock((GMutex *) mutex->once.retval);
-}
-#undef g_mutex_unlock
-
-static inline gpointer do_g_cond_new(gpointer unused)
-{
- return (gpointer) g_cond_new();
-}
-
-static inline void g_cond_init(CompatGCond *cond)
-{
- cond->once = (GOnce) G_ONCE_INIT;
-}
-
-static inline void g_cond_clear(CompatGCond *cond)
-{
- g_assert(cond->once.status != G_ONCE_STATUS_PROGRESS);
- if (cond->once.retval) {
- g_cond_free((GCond *) cond->once.retval);
- }
- cond->once = (GOnce) G_ONCE_INIT;
-}
-
-static inline void (g_cond_wait)(CompatGCond *cond, CompatGMutex *mutex)
-{
- g_assert(mutex->once.status != G_ONCE_STATUS_PROGRESS);
- g_once(&cond->once, do_g_cond_new, NULL);
- g_cond_wait((GCond *) cond->once.retval, (GMutex *) mutex->once.retval);
-}
-#undef g_cond_wait
-
-static inline void (g_cond_broadcast)(CompatGCond *cond)
-{
- g_once(&cond->once, do_g_cond_new, NULL);
- g_cond_broadcast((GCond *) cond->once.retval);
-}
-#undef g_cond_broadcast
-
-static inline void (g_cond_signal)(CompatGCond *cond)
-{
- g_once(&cond->once, do_g_cond_new, NULL);
- g_cond_signal((GCond *) cond->once.retval);
-}
-#undef g_cond_signal
-
-static inline gboolean (g_cond_timed_wait)(CompatGCond *cond,
- CompatGMutex *mutex,
- GTimeVal *time)
-{
- g_assert(mutex->once.status != G_ONCE_STATUS_PROGRESS);
- g_once(&cond->once, do_g_cond_new, NULL);
- return g_cond_timed_wait((GCond *) cond->once.retval,
- (GMutex *) mutex->once.retval, time);
-}
-#undef g_cond_timed_wait
-
-/* This is not a macro, because it didn't exist until 2.32. */
-static inline gboolean g_cond_wait_until(CompatGCond *cond, CompatGMutex *mutex,
- gint64 end_time)
-{
- GTimeVal time;
-
- /* Convert from monotonic to CLOCK_REALTIME. */
- end_time -= g_get_monotonic_time();
- g_get_current_time(&time);
- end_time += time.tv_sec * G_TIME_SPAN_SECOND + time.tv_usec;
-
- time.tv_sec = end_time / G_TIME_SPAN_SECOND;
- time.tv_usec = end_time % G_TIME_SPAN_SECOND;
- return g_cond_timed_wait(cond, mutex, &time);
-}
-
-/* before 2.31 there was no g_thread_new() */
-static inline GThread *g_thread_new(const char *name,
- GThreadFunc func, gpointer data)
-{
- GThread *thread = g_thread_create(func, data, TRUE, NULL);
- if (!thread) {
- g_error("creating thread");
- }
- return thread;
-}
-#else
-#define CompatGMutex GMutex
-#define CompatGCond GCond
-#endif /* glib 2.31 */
-
-#if !GLIB_CHECK_VERSION(2, 32, 0)
-/* Beware, function returns gboolean since 2.39.2, see GLib commit 9101915 */
-static inline void g_hash_table_add(GHashTable *hash_table, gpointer key)
-{
- g_hash_table_replace(hash_table, key, key);
-}
-
-static inline gboolean g_hash_table_contains(GHashTable *hash_table,
- gpointer key)
-{
- return g_hash_table_lookup_extended(hash_table, key, NULL, NULL);
-}
-#define G_SOURCE_CONTINUE TRUE
-#define G_SOURCE_REMOVE FALSE
-#endif
-
-#ifndef g_assert_true
-#define g_assert_true(expr) \
- do { \
- if (G_LIKELY(expr)) { \
- } else { \
- g_assertion_message(G_LOG_DOMAIN, __FILE__, __LINE__, G_STRFUNC, \
- "'" #expr "' should be TRUE"); \
- } \
- } while (0)
-#endif
-
-#ifndef g_assert_false
-#define g_assert_false(expr) \
- do { \
- if (G_LIKELY(!(expr))) { \
- } else { \
- g_assertion_message(G_LOG_DOMAIN, __FILE__, __LINE__, G_STRFUNC, \
- "'" #expr "' should be FALSE"); \
- } \
- } while (0)
-#endif
-
-#ifndef g_assert_null
-#define g_assert_null(expr) \
- do { \
- if (G_LIKELY((expr) == NULL)) { \
- } else { \
- g_assertion_message(G_LOG_DOMAIN, __FILE__, __LINE__, G_STRFUNC, \
- "'" #expr "' should be NULL"); \
- } \
- } while (0)
-#endif
-
-#ifndef g_assert_nonnull
-#define g_assert_nonnull(expr) \
- do { \
- if (G_LIKELY((expr) != NULL)) { \
- } else { \
- g_assertion_message(G_LOG_DOMAIN, __FILE__, __LINE__, G_STRFUNC, \
- "'" #expr "' should not be NULL"); \
- } \
- } while (0)
-#endif
#ifndef g_assert_cmpmem
#define g_assert_cmpmem(m1, l1, m2, l2) \
@@ -288,80 +45,4 @@ static inline gboolean g_hash_table_contains(GHashTable *hash_table,
} while (0)
#endif
-#if !GLIB_CHECK_VERSION(2, 28, 0)
-static inline void g_list_free_full(GList *list, GDestroyNotify free_func)
-{
- GList *l;
-
- for (l = list; l; l = l->next) {
- free_func(l->data);
- }
-
- g_list_free(list);
-}
-
-static inline void g_slist_free_full(GSList *list, GDestroyNotify free_func)
-{
- GSList *l;
-
- for (l = list; l; l = l->next) {
- free_func(l->data);
- }
-
- g_slist_free(list);
-}
-#endif
-
-#if !GLIB_CHECK_VERSION(2, 26, 0)
-static inline void g_source_set_name(GSource *source, const char *name)
-{
- /* This is just a debugging aid, so leaving it a no-op */
-}
-static inline void g_source_set_name_by_id(guint tag, const char *name)
-{
- /* This is just a debugging aid, so leaving it a no-op */
-}
-#endif
-
-#if !GLIB_CHECK_VERSION(2, 36, 0)
-/* Always fail. This will not include error_report output in the test log,
- * sending it instead to stderr.
- */
-#define g_test_initialized() (0)
-#endif
-#if !GLIB_CHECK_VERSION(2, 38, 0)
-#ifdef CONFIG_HAS_GLIB_SUBPROCESS_TESTS
-#error schizophrenic detection of glib subprocess testing
-#endif
-#define g_test_subprocess() (0)
-#endif
-
-
-#if !GLIB_CHECK_VERSION(2, 34, 0)
-static inline void
-g_test_add_data_func_full(const char *path,
- gpointer data,
- gpointer fn,
- gpointer data_free_func)
-{
-#if GLIB_CHECK_VERSION(2, 26, 0)
- /* back-compat casts, remove this once we can require new-enough glib */
- g_test_add_vtable(path, 0, data, NULL,
- (GTestFixtureFunc)fn, (GTestFixtureFunc) data_free_func);
-#else
- /* back-compat casts, remove this once we can require new-enough glib */
- g_test_add_vtable(path, 0, data, NULL,
- (void (*)(void)) fn, (void (*)(void)) data_free_func);
-#endif
-}
-#endif
-
-/* Small compat shim from glib 2.32 */
-#ifndef G_SOURCE_CONTINUE
-#define G_SOURCE_CONTINUE TRUE
-#endif
-#ifndef G_SOURCE_REMOVE
-#define G_SOURCE_REMOVE FALSE
-#endif
-
#endif
diff --git a/tests/docker/dockerfiles/centos6.docker b/tests/docker/dockerfiles/centos6.docker
deleted file mode 100644
index ad24319..0000000
--- a/tests/docker/dockerfiles/centos6.docker
+++ /dev/null
@@ -1,30 +0,0 @@
-FROM centos:6
-RUN yum install -y epel-release centos-release-xen
-ENV PACKAGES \
- bison \
- bzip2-devel \
- ccache \
- csnappy-devel \
- flex \
- g++ \
- gcc \
- gettext \
- git \
- glib2-devel \
- libepoxy-devel \
- libfdt-devel \
- librdmacm-devel \
- lzo-devel \
- make \
- mesa-libEGL-devel \
- mesa-libgbm-devel \
- pixman-devel \
- SDL-devel \
- spice-glib-devel \
- spice-server-devel \
- tar \
- vte-devel \
- xen-devel \
- zlib-devel
-RUN yum install -y $PACKAGES
-RUN rpm -q $PACKAGES | sort > /packages.txt
diff --git a/tests/docker/dockerfiles/min-glib.docker b/tests/docker/dockerfiles/min-glib.docker
deleted file mode 100644
index f2eed97..0000000
--- a/tests/docker/dockerfiles/min-glib.docker
+++ /dev/null
@@ -1,8 +0,0 @@
-FROM centos:6
-RUN yum install -y \
- tar gettext git make gcc g++ \
- zlib-devel SDL-devel pixman-devel \
- epel-release
-RUN yum install -y libfdt-devel ccache
-RUN yum downgrade -y http://vault.centos.org/6.0/os/x86_64/Packages/glib2-2.22.5-5.el6.x86_64.rpm
-RUN yum install -y http://vault.centos.org/6.0/os/x86_64/Packages/glib2-devel-2.22.5-5.el6.x86_64.rpm
diff --git a/tests/test-qmp-event.c b/tests/test-qmp-event.c
index 3a7c227..78a2942 100644
--- a/tests/test-qmp-event.c
+++ b/tests/test-qmp-event.c
@@ -32,7 +32,7 @@ typedef struct QDictCmpData {
} QDictCmpData;
TestEventData *test_event_data;
-static CompatGMutex test_event_lock;
+static GMutex test_event_lock;
/* Only compares bool, int, string */
static
diff --git a/tests/tpm-emu.h b/tests/tpm-emu.h
index ef4bfa8..08f9024 100644
--- a/tests/tpm-emu.h
+++ b/tests/tpm-emu.h
@@ -24,8 +24,8 @@ struct tpm_hdr {
} QEMU_PACKED;
typedef struct TestState {
- CompatGMutex data_mutex;
- CompatGCond data_cond;
+ GMutex data_mutex;
+ GCond data_cond;
SocketAddress *addr;
QIOChannel *tpm_ioc;
GThread *emu_tpm_thread;
diff --git a/tests/vhost-user-test.c b/tests/vhost-user-test.c
index bbc8091..9f680f5 100644
--- a/tests/vhost-user-test.c
+++ b/tests/vhost-user-test.c
@@ -150,8 +150,8 @@ typedef struct TestServer {
int fds_num;
int fds[VHOST_MEMORY_MAX_NREGIONS];
VhostUserMemory memory;
- CompatGMutex data_mutex;
- CompatGCond data_cond;
+ GMutex data_mutex;
+ GCond data_cond;
int log_fd;
uint64_t rings;
bool test_fail;
diff --git a/trace/simple.c b/trace/simple.c
index e82018d..701dec6 100644
--- a/trace/simple.c
+++ b/trace/simple.c
@@ -36,9 +36,9 @@
* Trace records are written out by a dedicated thread. The thread waits for
* records to become available, writes them out, and then waits again.
*/
-static CompatGMutex trace_lock;
-static CompatGCond trace_available_cond;
-static CompatGCond trace_empty_cond;
+static GMutex trace_lock;
+static GCond trace_available_cond;
+static GCond trace_empty_cond;
static bool trace_available;
static bool trace_writeout_enabled;
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* [Qemu-devel] [PULL 24/30] i386/kvm: add support for Hyper-V reenlightenment MSRs
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
` (22 preceding siblings ...)
2018-05-08 22:14 ` [Qemu-devel] [PULL 23/30] glib: bump min required glib library version to 2.42 Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 25/30] configure: Really use local libfdt if the system one is too old Paolo Bonzini
` (6 subsequent siblings)
30 siblings, 0 replies; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel; +Cc: Vitaly Kuznetsov
From: Vitaly Kuznetsov <vkuznets@redhat.com>
KVM recently gained support for Hyper-V Reenlightenment MSRs which are
required to make KVM-on-Hyper-V enable TSC page clocksource to its guests
when INVTSC is not passed to it (and it is not passed by default in Qemu
as it effectively blocks migration).
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Message-Id: <20180411115036.31832-2-vkuznets@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
target/i386/cpu.c | 4 +++-
target/i386/cpu.h | 4 ++++
target/i386/hyperv-proto.h | 9 ++++++++-
target/i386/kvm.c | 39 ++++++++++++++++++++++++++++++++++++++-
target/i386/machine.c | 24 ++++++++++++++++++++++++
5 files changed, 77 insertions(+), 3 deletions(-)
diff --git a/target/i386/cpu.c b/target/i386/cpu.c
index a20fe26..b0a1c62 100644
--- a/target/i386/cpu.c
+++ b/target/i386/cpu.c
@@ -416,7 +416,8 @@ static FeatureWordInfo feature_word_info[FEATURE_WORDS] = {
NULL /* hv_vpindex_access */, NULL /* hv_msr_reset_access */,
NULL /* hv_msr_stats_access */, NULL /* hv_reftsc_access */,
NULL /* hv_msr_idle_access */, NULL /* hv_msr_frequency_access */,
- NULL, NULL, NULL, NULL,
+ NULL /* hv_msr_debug_access */, NULL /* hv_msr_reenlightenment_access */,
+ NULL, NULL,
NULL, NULL, NULL, NULL,
NULL, NULL, NULL, NULL,
NULL, NULL, NULL, NULL,
@@ -4770,6 +4771,7 @@ static Property x86_cpu_properties[] = {
DEFINE_PROP_BOOL("hv-synic", X86CPU, hyperv_synic, false),
DEFINE_PROP_BOOL("hv-stimer", X86CPU, hyperv_stimer, false),
DEFINE_PROP_BOOL("hv-frequencies", X86CPU, hyperv_frequencies, false),
+ DEFINE_PROP_BOOL("hv-reenlightenment", X86CPU, hyperv_reenlightenment, false),
DEFINE_PROP_BOOL("check", X86CPU, check_cpuid, true),
DEFINE_PROP_BOOL("enforce", X86CPU, enforce_cpuid, false),
DEFINE_PROP_BOOL("kvm", X86CPU, expose_kvm, true),
diff --git a/target/i386/cpu.h b/target/i386/cpu.h
index 1b219fa..b58b779 100644
--- a/target/i386/cpu.h
+++ b/target/i386/cpu.h
@@ -1174,6 +1174,9 @@ typedef struct CPUX86State {
uint64_t msr_hv_synic_sint[HV_SINT_COUNT];
uint64_t msr_hv_stimer_config[HV_STIMER_COUNT];
uint64_t msr_hv_stimer_count[HV_STIMER_COUNT];
+ uint64_t msr_hv_reenlightenment_control;
+ uint64_t msr_hv_tsc_emulation_control;
+ uint64_t msr_hv_tsc_emulation_status;
uint64_t msr_rtit_ctrl;
uint64_t msr_rtit_status;
@@ -1297,6 +1300,7 @@ struct X86CPU {
bool hyperv_synic;
bool hyperv_stimer;
bool hyperv_frequencies;
+ bool hyperv_reenlightenment;
bool check_cpuid;
bool enforce_cpuid;
bool expose_kvm;
diff --git a/target/i386/hyperv-proto.h b/target/i386/hyperv-proto.h
index cb4d7f2..93352eb 100644
--- a/target/i386/hyperv-proto.h
+++ b/target/i386/hyperv-proto.h
@@ -35,7 +35,7 @@
#define HV_RESET_AVAILABLE (1u << 7)
#define HV_REFERENCE_TSC_AVAILABLE (1u << 9)
#define HV_ACCESS_FREQUENCY_MSRS (1u << 11)
-
+#define HV_ACCESS_REENLIGHTENMENTS_CONTROL (1u << 13)
/*
* HV_CPUID_FEATURES.EDX bits
@@ -130,6 +130,13 @@
#define HV_CRASH_CTL_NOTIFY (1ull << 63)
/*
+ * Reenlightenment notification MSRs
+ */
+#define HV_X64_MSR_REENLIGHTENMENT_CONTROL 0x40000106
+#define HV_X64_MSR_TSC_EMULATION_CONTROL 0x40000107
+#define HV_X64_MSR_TSC_EMULATION_STATUS 0x40000108
+
+/*
* Hypercall status code
*/
#define HV_STATUS_SUCCESS 0
diff --git a/target/i386/kvm.c b/target/i386/kvm.c
index 6c49954..d6666a4 100644
--- a/target/i386/kvm.c
+++ b/target/i386/kvm.c
@@ -90,6 +90,7 @@ static bool has_msr_hv_runtime;
static bool has_msr_hv_synic;
static bool has_msr_hv_stimer;
static bool has_msr_hv_frequencies;
+static bool has_msr_hv_reenlightenment;
static bool has_msr_xss;
static bool has_msr_spec_ctrl;
static bool has_msr_smi_count;
@@ -583,7 +584,8 @@ static bool hyperv_enabled(X86CPU *cpu)
cpu->hyperv_vpindex ||
cpu->hyperv_runtime ||
cpu->hyperv_synic ||
- cpu->hyperv_stimer);
+ cpu->hyperv_stimer ||
+ cpu->hyperv_reenlightenment);
}
static int kvm_arch_set_tsc_khz(CPUState *cs)
@@ -669,6 +671,16 @@ static int hyperv_handle_properties(CPUState *cs)
}
env->features[FEAT_HYPERV_EDX] |= HV_GUEST_CRASH_MSR_AVAILABLE;
}
+ if (cpu->hyperv_reenlightenment) {
+ if (!has_msr_hv_reenlightenment) {
+ fprintf(stderr,
+ "Hyper-V Reenlightenment MSRs "
+ "(requested by 'hv-reenlightenment' cpu flag) "
+ "are not supported by kernel\n");
+ return -ENOSYS;
+ }
+ env->features[FEAT_HYPERV_EAX] |= HV_ACCESS_REENLIGHTENMENTS_CONTROL;
+ }
env->features[FEAT_HYPERV_EDX] |= HV_CPU_DYNAMIC_PARTITIONING_AVAILABLE;
if (cpu->hyperv_reset) {
if (!has_msr_hv_reset) {
@@ -1215,6 +1227,9 @@ static int kvm_get_supported_msrs(KVMState *s)
case HV_X64_MSR_TSC_FREQUENCY:
has_msr_hv_frequencies = true;
break;
+ case HV_X64_MSR_REENLIGHTENMENT_CONTROL:
+ has_msr_hv_reenlightenment = true;
+ break;
case MSR_IA32_SPEC_CTRL:
has_msr_spec_ctrl = true;
break;
@@ -1778,6 +1793,14 @@ static int kvm_put_msrs(X86CPU *cpu, int level)
kvm_msr_entry_add(cpu, HV_X64_MSR_REFERENCE_TSC,
env->msr_hv_tsc);
}
+ if (cpu->hyperv_reenlightenment) {
+ kvm_msr_entry_add(cpu, HV_X64_MSR_REENLIGHTENMENT_CONTROL,
+ env->msr_hv_reenlightenment_control);
+ kvm_msr_entry_add(cpu, HV_X64_MSR_TSC_EMULATION_CONTROL,
+ env->msr_hv_tsc_emulation_control);
+ kvm_msr_entry_add(cpu, HV_X64_MSR_TSC_EMULATION_STATUS,
+ env->msr_hv_tsc_emulation_status);
+ }
}
if (cpu->hyperv_vapic) {
kvm_msr_entry_add(cpu, HV_X64_MSR_APIC_ASSIST_PAGE,
@@ -2140,6 +2163,11 @@ static int kvm_get_msrs(X86CPU *cpu)
if (cpu->hyperv_time) {
kvm_msr_entry_add(cpu, HV_X64_MSR_REFERENCE_TSC, 0);
}
+ if (cpu->hyperv_reenlightenment) {
+ kvm_msr_entry_add(cpu, HV_X64_MSR_REENLIGHTENMENT_CONTROL, 0);
+ kvm_msr_entry_add(cpu, HV_X64_MSR_TSC_EMULATION_CONTROL, 0);
+ kvm_msr_entry_add(cpu, HV_X64_MSR_TSC_EMULATION_STATUS, 0);
+ }
if (has_msr_hv_crash) {
int j;
@@ -2397,6 +2425,15 @@ static int kvm_get_msrs(X86CPU *cpu)
env->msr_hv_stimer_count[(index - HV_X64_MSR_STIMER0_COUNT)/2] =
msrs[i].data;
break;
+ case HV_X64_MSR_REENLIGHTENMENT_CONTROL:
+ env->msr_hv_reenlightenment_control = msrs[i].data;
+ break;
+ case HV_X64_MSR_TSC_EMULATION_CONTROL:
+ env->msr_hv_tsc_emulation_control = msrs[i].data;
+ break;
+ case HV_X64_MSR_TSC_EMULATION_STATUS:
+ env->msr_hv_tsc_emulation_status = msrs[i].data;
+ break;
case MSR_MTRRdefType:
env->mtrr_deftype = msrs[i].data;
break;
diff --git a/target/i386/machine.c b/target/i386/machine.c
index bd2d82e..fd99c0b 100644
--- a/target/i386/machine.c
+++ b/target/i386/machine.c
@@ -713,6 +713,29 @@ static const VMStateDescription vmstate_msr_hyperv_stimer = {
}
};
+static bool hyperv_reenlightenment_enable_needed(void *opaque)
+{
+ X86CPU *cpu = opaque;
+ CPUX86State *env = &cpu->env;
+
+ return env->msr_hv_reenlightenment_control != 0 ||
+ env->msr_hv_tsc_emulation_control != 0 ||
+ env->msr_hv_tsc_emulation_status != 0;
+}
+
+static const VMStateDescription vmstate_msr_hyperv_reenlightenment = {
+ .name = "cpu/msr_hyperv_reenlightenment",
+ .version_id = 1,
+ .minimum_version_id = 1,
+ .needed = hyperv_reenlightenment_enable_needed,
+ .fields = (VMStateField[]) {
+ VMSTATE_UINT64(env.msr_hv_reenlightenment_control, X86CPU),
+ VMSTATE_UINT64(env.msr_hv_tsc_emulation_control, X86CPU),
+ VMSTATE_UINT64(env.msr_hv_tsc_emulation_status, X86CPU),
+ VMSTATE_END_OF_LIST()
+ }
+};
+
static bool avx512_needed(void *opaque)
{
X86CPU *cpu = opaque;
@@ -1005,6 +1028,7 @@ VMStateDescription vmstate_x86_cpu = {
&vmstate_msr_hyperv_runtime,
&vmstate_msr_hyperv_synic,
&vmstate_msr_hyperv_stimer,
+ &vmstate_msr_hyperv_reenlightenment,
&vmstate_avx512,
&vmstate_xss,
&vmstate_tsc_khz,
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* [Qemu-devel] [PULL 25/30] configure: Really use local libfdt if the system one is too old
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
` (23 preceding siblings ...)
2018-05-08 22:14 ` [Qemu-devel] [PULL 24/30] i386/kvm: add support for Hyper-V reenlightenment MSRs Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 26/30] configure: Display if libfdt is from system or git Paolo Bonzini
` (5 subsequent siblings)
30 siblings, 0 replies; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel; +Cc: Philippe Mathieu-Daudé
From: Philippe Mathieu-Daudé <f4bug@amsat.org>
QEMU requires libfdt version >= 1.4.2.
If the host has an older libfdt installed, the configure script will use
a (git cloned) local version.
Example with Debian 8:
$ dpkg-query --showformat='${Version}\n' --show libfdt-dev
1.4.0+dfsg-1
$ ./configure
[...]
fdt support yes # from git submodule 'dtc'
If this case occurs, the linker will have 2 different libfdt available in
the library search path. The default behavior is to search the system path
first, then the local path.
Even if the configure script noticed the libfdt is too old and clone a more
recent locally, when linking the system library is selected first, and the
link process eventually fails:
LINK mips64el-softmmu/qemu-system-mips64el
../hw/core/loader-fit.o: In function `load_fit':
/root/src/github.com/philmd/qemu/hw/core/loader-fit.c:278: undefined reference to `fdt_first_subnode'
/root/src/github.com/philmd/qemu/hw/core/loader-fit.c:286: undefined reference to `fdt_next_subnode'
/root/src/github.com/philmd/qemu/hw/core/loader-fit.c:277: undefined reference to `fdt_first_subnode'
collect2: error: ld returned 1 exit status
Makefile:201: recipe for target 'qemu-system-mips64el' failed
make[1]: *** [qemu-system-mips64el] Error 1
QEMU already uses a kludge to enforce local CFLAGS before system ones for
libpixman and libfdt, add a similar kludge for the LDFLAGS to enforce using
the local libfdt.
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20180415230522.24404-2-f4bug@amsat.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
---
configure | 6 +++++-
rules.mak | 2 +-
2 files changed, 6 insertions(+), 2 deletions(-)
diff --git a/configure b/configure
index 3ff4d77..43756dc 100755
--- a/configure
+++ b/configure
@@ -3800,7 +3800,8 @@ EOF
symlink "$source_path/dtc/scripts" "dtc/scripts"
fi
fdt_cflags="-I\$(SRC_PATH)/dtc/libfdt"
- fdt_libs="-L\$(BUILD_DIR)/dtc/libfdt $fdt_libs"
+ fdt_ldflags="-L\$(BUILD_DIR)/dtc/libfdt"
+ fdt_libs="$fdt_libs"
elif test "$fdt" = "yes" ; then
# Not a git build & no libfdt found, prompt for system install
error_exit "DTC (libfdt) version >= 1.4.2 not present." \
@@ -5742,6 +5743,7 @@ echo_version() {
# prepend pixman and ftd flags after all config tests are done
QEMU_CFLAGS="$pixman_cflags $fdt_cflags $QEMU_CFLAGS"
+QEMU_LDFLAGS="$fdt_ldflags $QEMU_LDFLAGS"
libs_softmmu="$pixman_libs $libs_softmmu"
echo "Install prefix $prefix"
@@ -5772,6 +5774,7 @@ echo "ARFLAGS $ARFLAGS"
echo "CFLAGS $CFLAGS"
echo "QEMU_CFLAGS $QEMU_CFLAGS"
echo "LDFLAGS $LDFLAGS"
+echo "QEMU_LDFLAGS $QEMU_LDFLAGS"
echo "make $make"
echo "install $install"
echo "python $python"
@@ -6706,6 +6709,7 @@ else
fi
echo "LDFLAGS=$LDFLAGS" >> $config_host_mak
echo "LDFLAGS_NOPIE=$LDFLAGS_NOPIE" >> $config_host_mak
+echo "QEMU_LDFLAGS=$QEMU_LDFLAGS" >> $config_host_mak
echo "LD_REL_FLAGS=$LD_REL_FLAGS" >> $config_host_mak
echo "LD_I386_EMULATION=$ld_i386_emulation" >> $config_host_mak
echo "LIBS+=$LIBS" >> $config_host_mak
diff --git a/rules.mak b/rules.mak
index 93a0702..04c7f74 100644
--- a/rules.mak
+++ b/rules.mak
@@ -73,7 +73,7 @@ expand-objs = $(strip $(sort $(filter %.o,$1)) \
# must link with the C++ compiler, not the plain C compiler.
LINKPROG = $(or $(CXX),$(CC))
-LINK = $(call quiet-command, $(LINKPROG) $(QEMU_CFLAGS) $(CFLAGS) $(LDFLAGS) -o $@ \
+LINK = $(call quiet-command, $(LINKPROG) $(QEMU_LDFLAGS) $(QEMU_CFLAGS) $(CFLAGS) $(LDFLAGS) -o $@ \
$(call process-archive-undefs, $1) \
$(version-obj-y) $(call extract-libs,$1) $(LIBS),"LINK","$(TARGET_DIR)$@")
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* [Qemu-devel] [PULL 26/30] configure: Display if libfdt is from system or git
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
` (24 preceding siblings ...)
2018-05-08 22:14 ` [Qemu-devel] [PULL 25/30] configure: Really use local libfdt if the system one is too old Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 27/30] shippable: Remove Debian 8 libfdt kludge Paolo Bonzini
` (4 subsequent siblings)
30 siblings, 0 replies; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel; +Cc: Philippe Mathieu-Daudé
From: Philippe Mathieu-Daudé <f4bug@amsat.org>
The configure script outputs "yes" regardless which libfdt is used:
./configure
[...]
fdt support yes
Sometimes you can have both system and local git version available,
change the configure script to display which library got selected:
debian8$ dpkg-query --showformat='${Version}\n' --show libfdt-dev
1.4.0+dfsg-1
debian8$ ./configure
[...]
fdt support git
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20180415230522.24404-3-f4bug@amsat.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
configure | 9 ++++-----
1 file changed, 4 insertions(+), 5 deletions(-)
diff --git a/configure b/configure
index 43756dc..407bf84 100755
--- a/configure
+++ b/configure
@@ -3785,15 +3785,14 @@ int main(void) { fdt_first_subnode(0, 0); return 0; }
EOF
if compile_prog "" "$fdt_libs" ; then
# system DTC is good - use it
- fdt=yes
+ fdt=system
else
# have GIT checkout, so activate dtc submodule
if test -e "${source_path}/.git" ; then
git_submodules="${git_submodules} dtc"
fi
if test -d "${source_path}/dtc/libfdt" || test -e "${source_path}/.git" ; then
- fdt=yes
- dtc_internal="yes"
+ fdt=git
mkdir -p dtc
if [ "$pwd_is_source_path" != "y" ] ; then
symlink "$source_path/dtc/Makefile" "dtc/Makefile"
@@ -6334,7 +6333,7 @@ fi
if test "$preadv" = "yes" ; then
echo "CONFIG_PREADV=y" >> $config_host_mak
fi
-if test "$fdt" = "yes" ; then
+if test "$fdt" != "no" ; then
echo "CONFIG_FDT=y" >> $config_host_mak
fi
if test "$membarrier" = "yes" ; then
@@ -7136,7 +7135,7 @@ echo "QEMU_CFLAGS+=$cflags" >> $config_target_mak
done # for target in $targets
-if [ "$dtc_internal" = "yes" ]; then
+if [ "$fdt" = "git" ]; then
echo "config-host.h: subdir-dtc" >> $config_host_mak
fi
if [ "$capstone" = "git" -o "$capstone" = "internal" ]; then
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* [Qemu-devel] [PULL 27/30] shippable: Remove Debian 8 libfdt kludge
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
` (25 preceding siblings ...)
2018-05-08 22:14 ` [Qemu-devel] [PULL 26/30] configure: Display if libfdt is from system or git Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 28/30] build: Silence dtc directory creation Paolo Bonzini
` (3 subsequent siblings)
30 siblings, 0 replies; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel; +Cc: Philippe Mathieu-Daudé
From: Philippe Mathieu-Daudé <f4bug@amsat.org>
This kludge was added in a825ca06137, but a cleaner and more generic
fix is now available (see ##COMMIT_CONFIGURE_LIBFDT_LDFLAGS_SHA##).
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20180415230522.24404-4-f4bug@amsat.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
.shippable.yml | 8 --------
1 file changed, 8 deletions(-)
diff --git a/.shippable.yml b/.shippable.yml
index 60f2ce9..f74a3de 100644
--- a/.shippable.yml
+++ b/.shippable.yml
@@ -35,13 +35,5 @@ build:
options: "-e HOME=/root"
ci:
- unset CC
- # some targets require newer up to date packages, for example TARGET_LIST matching
- # aarch64*-softmmu|arm*-softmmu|ppc*-softmmu|microblaze*-softmmu|mips64el-softmmu)
- # see the configure script:
- # error_exit "DTC (libfdt) version >= 1.4.2 not present. Your options:"
- # " (1) Preferred: Install the DTC (libfdt) devel package"
- # " (2) Fetch the DTC submodule, using:"
- # " git submodule update --init dtc"
- - dpkg --compare-versions `dpkg-query --showformat='${Version}' --show libfdt-dev` ge 1.4.2 || git submodule update --init dtc
- ./configure ${QEMU_CONFIGURE_OPTS} --target-list=${TARGET_LIST}
- make -j$(($(getconf _NPROCESSORS_ONLN) + 1))
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* [Qemu-devel] [PULL 28/30] build: Silence dtc directory creation
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
` (26 preceding siblings ...)
2018-05-08 22:14 ` [Qemu-devel] [PULL 27/30] shippable: Remove Debian 8 libfdt kludge Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 29/30] pc-dimm: fix error messages if no slots were defined Paolo Bonzini
` (2 subsequent siblings)
30 siblings, 0 replies; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel; +Cc: Jan Kiszka
From: Jan Kiszka <jan.kiszka@siemens.com>
Align with other mkdir calls.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <0dd4c8f5-d60e-e564-652f-cd0101f6ee68@web.de>
Message-Id: <20180415230522.24404-5-f4bug@amsat.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
Makefile | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/Makefile b/Makefile
index d71dd5b..35554b5 100644
--- a/Makefile
+++ b/Makefile
@@ -485,7 +485,7 @@ subdir-dtc: .git-submodule-status dtc/libfdt dtc/tests
$(call quiet-command,$(MAKE) $(DTC_MAKE_ARGS) CPPFLAGS="$(DTC_CPPFLAGS)" CFLAGS="$(DTC_CFLAGS)" LDFLAGS="$(LDFLAGS)" ARFLAGS="$(ARFLAGS)" CC="$(CC)" AR="$(AR)" LD="$(LD)" $(SUBDIR_MAKEFLAGS) libfdt/libfdt.a,)
dtc/%: .git-submodule-status
- mkdir -p $@
+ @mkdir -p $@
# Overriding CFLAGS causes us to lose defines added in the sub-makefile.
# Not overriding CFLAGS leads to mis-matches between compilation modes.
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* [Qemu-devel] [PULL 29/30] pc-dimm: fix error messages if no slots were defined
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
` (27 preceding siblings ...)
2018-05-08 22:14 ` [Qemu-devel] [PULL 28/30] build: Silence dtc directory creation Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-08 22:14 ` [Qemu-devel] [PULL 30/30] rename included C files to foo.inc.c, remove osdep.h Paolo Bonzini
2018-05-11 12:19 ` [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Peter Maydell
30 siblings, 0 replies; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel; +Cc: David Hildenbrand
From: David Hildenbrand <david@redhat.com>
If no slots were defined we try to allocate an empty bitmap, which
fails.
Signed-off-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Marcel Apfelbaum <marcel@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Message-Id: <20180427120515.24067-1-david@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
hw/mem/pc-dimm.c | 9 ++++++++-
1 file changed, 8 insertions(+), 1 deletion(-)
diff --git a/hw/mem/pc-dimm.c b/hw/mem/pc-dimm.c
index 0119c68..12da89d 100644
--- a/hw/mem/pc-dimm.c
+++ b/hw/mem/pc-dimm.c
@@ -118,9 +118,16 @@ static int pc_dimm_slot2bitmap(Object *obj, void *opaque)
int pc_dimm_get_free_slot(const int *hint, int max_slots, Error **errp)
{
- unsigned long *bitmap = bitmap_new(max_slots);
+ unsigned long *bitmap;
int slot = 0;
+ if (max_slots <= 0) {
+ error_setg(errp, "no slots where allocated, please specify "
+ "the 'slots' option");
+ return slot;
+ }
+
+ bitmap = bitmap_new(max_slots);
object_child_foreach(qdev_get_machine(), pc_dimm_slot2bitmap, bitmap);
/* check if requested slot is not occupied */
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* [Qemu-devel] [PULL 30/30] rename included C files to foo.inc.c, remove osdep.h
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
` (28 preceding siblings ...)
2018-05-08 22:14 ` [Qemu-devel] [PULL 29/30] pc-dimm: fix error messages if no slots were defined Paolo Bonzini
@ 2018-05-08 22:14 ` Paolo Bonzini
2018-05-11 12:19 ` [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Peter Maydell
30 siblings, 0 replies; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-08 22:14 UTC (permalink / raw)
To: qemu-devel; +Cc: Eric Blake, Michael S. Tsirkin
osdep.h is only needed for files that are compiled directly.
Remove it from included C source files, and rename them to
*.inc.c so that scripts/clean-includes knows to skip them.
Cc: Eric Blake <eblake@redhat.com>
Cc: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
target/cris/translate.c | 2 +-
.../cris/{translate_v10.c => translate_v10.inc.c} | 0
target/mips/translate.c | 2 +-
.../{translate_init.c => translate_init.inc.c} | 0
target/ppc/int_helper.c | 2 +-
target/ppc/{mfrom_table.c => mfrom_table.inc.c} | 0
target/ppc/translate.c | 2 +-
.../ppc/{translate_init.c => translate_init.inc.c} | 1 -
ui/vnc-enc-zrle.c | 22 +++++++++++-----------
ui/{vnc-enc-zrle-template.c => vnc-enc-zrle.inc.c} | 0
10 files changed, 15 insertions(+), 16 deletions(-)
rename target/cris/{translate_v10.c => translate_v10.inc.c} (100%)
rename target/mips/{translate_init.c => translate_init.inc.c} (100%)
rename target/ppc/{mfrom_table.c => mfrom_table.inc.c} (100%)
rename target/ppc/{translate_init.c => translate_init.inc.c} (99%)
rename ui/{vnc-enc-zrle-template.c => vnc-enc-zrle.inc.c} (100%)
diff --git a/target/cris/translate.c b/target/cris/translate.c
index f51a731..ce1ee7a 100644
--- a/target/cris/translate.c
+++ b/target/cris/translate.c
@@ -3047,7 +3047,7 @@ static unsigned int crisv32_decoder(CPUCRISState *env, DisasContext *dc)
return insn_len;
}
-#include "translate_v10.c"
+#include "translate_v10.inc.c"
/*
* Delay slots on QEMU/CRIS.
diff --git a/target/cris/translate_v10.c b/target/cris/translate_v10.inc.c
similarity index 100%
rename from target/cris/translate_v10.c
rename to target/cris/translate_v10.inc.c
diff --git a/target/mips/translate.c b/target/mips/translate.c
index d05ee67..26f5404 100644
--- a/target/mips/translate.c
+++ b/target/mips/translate.c
@@ -20499,7 +20499,7 @@ void mips_tcg_init(void)
"fcr31");
}
-#include "translate_init.c"
+#include "translate_init.inc.c"
void cpu_mips_realize_env(CPUMIPSState *env)
{
diff --git a/target/mips/translate_init.c b/target/mips/translate_init.inc.c
similarity index 100%
rename from target/mips/translate_init.c
rename to target/mips/translate_init.inc.c
diff --git a/target/ppc/int_helper.c b/target/ppc/int_helper.c
index 35bdf09..1607a7a 100644
--- a/target/ppc/int_helper.c
+++ b/target/ppc/int_helper.c
@@ -379,7 +379,7 @@ target_ulong helper_divso(CPUPPCState *env, target_ulong arg1,
target_ulong helper_602_mfrom(target_ulong arg)
{
if (likely(arg < 602)) {
-#include "mfrom_table.c"
+#include "mfrom_table.inc.c"
return mfrom_ROM_table[arg];
} else {
return 0;
diff --git a/target/ppc/mfrom_table.c b/target/ppc/mfrom_table.inc.c
similarity index 100%
rename from target/ppc/mfrom_table.c
rename to target/ppc/mfrom_table.inc.c
diff --git a/target/ppc/translate.c b/target/ppc/translate.c
index 2a4140f..257badf 100644
--- a/target/ppc/translate.c
+++ b/target/ppc/translate.c
@@ -6991,7 +6991,7 @@ GEN_HANDLER2_E(trechkpt, "trechkpt", 0x1F, 0x0E, 0x1F, 0x03FFF800, \
};
#include "helper_regs.h"
-#include "translate_init.c"
+#include "translate_init.inc.c"
/*****************************************************************************/
/* Misc PowerPC helpers */
diff --git a/target/ppc/translate_init.c b/target/ppc/translate_init.inc.c
similarity index 99%
rename from target/ppc/translate_init.c
rename to target/ppc/translate_init.inc.c
index a72be6d..a0b3f18 100644
--- a/target/ppc/translate_init.c
+++ b/target/ppc/translate_init.inc.c
@@ -18,7 +18,6 @@
* License along with this library; if not, see <http://www.gnu.org/licenses/>.
*/
-#include "qemu/osdep.h"
#include "disas/bfd.h"
#include "exec/gdbstub.h"
#include "kvm_ppc.h"
diff --git a/ui/vnc-enc-zrle.c b/ui/vnc-enc-zrle.c
index fd63d4f..7493a84 100644
--- a/ui/vnc-enc-zrle.c
+++ b/ui/vnc-enc-zrle.c
@@ -199,56 +199,56 @@ static void zrle_write_u8(VncState *vs, uint8_t value)
#define ZRLE_BPP 8
#define ZYWRLE_ENDIAN ENDIAN_NO
-#include "vnc-enc-zrle-template.c"
+#include "vnc-enc-zrle.inc.c"
#undef ZRLE_BPP
#define ZRLE_BPP 15
#undef ZYWRLE_ENDIAN
#define ZYWRLE_ENDIAN ENDIAN_LITTLE
-#include "vnc-enc-zrle-template.c"
+#include "vnc-enc-zrle.inc.c"
#undef ZYWRLE_ENDIAN
#define ZYWRLE_ENDIAN ENDIAN_BIG
-#include "vnc-enc-zrle-template.c"
+#include "vnc-enc-zrle.inc.c"
#undef ZRLE_BPP
#define ZRLE_BPP 16
#undef ZYWRLE_ENDIAN
#define ZYWRLE_ENDIAN ENDIAN_LITTLE
-#include "vnc-enc-zrle-template.c"
+#include "vnc-enc-zrle.inc.c"
#undef ZYWRLE_ENDIAN
#define ZYWRLE_ENDIAN ENDIAN_BIG
-#include "vnc-enc-zrle-template.c"
+#include "vnc-enc-zrle.inc.c"
#undef ZRLE_BPP
#define ZRLE_BPP 32
#undef ZYWRLE_ENDIAN
#define ZYWRLE_ENDIAN ENDIAN_LITTLE
-#include "vnc-enc-zrle-template.c"
+#include "vnc-enc-zrle.inc.c"
#undef ZYWRLE_ENDIAN
#define ZYWRLE_ENDIAN ENDIAN_BIG
-#include "vnc-enc-zrle-template.c"
+#include "vnc-enc-zrle.inc.c"
#define ZRLE_COMPACT_PIXEL 24a
#undef ZYWRLE_ENDIAN
#define ZYWRLE_ENDIAN ENDIAN_LITTLE
-#include "vnc-enc-zrle-template.c"
+#include "vnc-enc-zrle.inc.c"
#undef ZYWRLE_ENDIAN
#define ZYWRLE_ENDIAN ENDIAN_BIG
-#include "vnc-enc-zrle-template.c"
+#include "vnc-enc-zrle.inc.c"
#undef ZRLE_COMPACT_PIXEL
#define ZRLE_COMPACT_PIXEL 24b
#undef ZYWRLE_ENDIAN
#define ZYWRLE_ENDIAN ENDIAN_LITTLE
-#include "vnc-enc-zrle-template.c"
+#include "vnc-enc-zrle.inc.c"
#undef ZYWRLE_ENDIAN
#define ZYWRLE_ENDIAN ENDIAN_BIG
-#include "vnc-enc-zrle-template.c"
+#include "vnc-enc-zrle.inc.c"
#undef ZRLE_COMPACT_PIXEL
#undef ZRLE_BPP
diff --git a/ui/vnc-enc-zrle-template.c b/ui/vnc-enc-zrle.inc.c
similarity index 100%
rename from ui/vnc-enc-zrle-template.c
rename to ui/vnc-enc-zrle.inc.c
--
1.8.3.1
^ permalink raw reply related [flat|nested] 40+ messages in thread* Re: [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09
2018-05-08 22:14 [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Paolo Bonzini
` (29 preceding siblings ...)
2018-05-08 22:14 ` [Qemu-devel] [PULL 30/30] rename included C files to foo.inc.c, remove osdep.h Paolo Bonzini
@ 2018-05-11 12:19 ` Peter Maydell
2018-05-11 12:33 ` Paolo Bonzini
2018-05-11 12:42 ` Daniel P. Berrangé
30 siblings, 2 replies; 40+ messages in thread
From: Peter Maydell @ 2018-05-11 12:19 UTC (permalink / raw)
To: Paolo Bonzini; +Cc: QEMU Developers
On 8 May 2018 at 23:14, Paolo Bonzini <pbonzini@redhat.com> wrote:
> The following changes since commit cc8f8ba754bba17eea9791d67b572eb26e30b4ce:
>
> Merge remote-tracking branch 'remotes/ehabkost/tags/machine-next-pull-request' into staging (2018-05-08 15:25:17 +0100)
>
> are available in the git repository at:
>
>
> git://github.com/bonzini/qemu.git tags/for-upstream
>
> for you to fetch changes up to 3c128df2ec239f5c1abd0aa59254c433a779a347:
>
> rename included C files to foo.inc.c, remove osdep.h (2018-05-09 00:13:41 +0200)
>
> ----------------------------------------------------------------
> * Don't silently truncate extremely long words in the command line
> * bump min required glib library version to 2.42
> * dtc configure fixes
> * MemoryRegionCache second try
> * Deprecated option removal
> * add support for Hyper-V reenlightenment MSRs
>
> ----------------------------------------------------------------
Some of my build setups barf on the new glib version requirement:
* my aarch64 build host (a gcc compile farm machine which is running
Ubuntu 14.04.5 LTS and has glib 2.40.2)
* my windows cross compile setups (which have glib 2.34.3)
The latter could probably be fixed, but it will be a pain because
gtk.org don't distribute prebuilt glib &c zipfiles usable for cross
compile any more. For the former, that LTS is apparently still
in support for another year.
thanks
-- PMM
^ permalink raw reply [flat|nested] 40+ messages in thread* Re: [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09
2018-05-11 12:19 ` [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Peter Maydell
@ 2018-05-11 12:33 ` Paolo Bonzini
2018-05-11 12:39 ` Peter Maydell
2018-05-11 12:42 ` Daniel P. Berrangé
1 sibling, 1 reply; 40+ messages in thread
From: Paolo Bonzini @ 2018-05-11 12:33 UTC (permalink / raw)
To: Peter Maydell; +Cc: QEMU Developers
On 11/05/2018 14:19, Peter Maydell wrote:
> Some of my build setups barf on the new glib version requirement:
>
> * my aarch64 build host (a gcc compile farm machine which is running
> Ubuntu 14.04.5 LTS and has glib 2.40.2)
> * my windows cross compile setups (which have glib 2.34.3)
>
> The latter could probably be fixed, but it will be a pain because
> gtk.org don't distribute prebuilt glib &c zipfiles usable for cross
> compile any more. For the former, that LTS is apparently still
> in support for another year.
Fair enough. I suppose that for now we can still uniformly bump up to
2.34, which is enough to remove all compatibility shims except for a couple.
I'll resend the pull request without that part.
Thanks,
Paolo
^ permalink raw reply [flat|nested] 40+ messages in thread
* Re: [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09
2018-05-11 12:33 ` Paolo Bonzini
@ 2018-05-11 12:39 ` Peter Maydell
0 siblings, 0 replies; 40+ messages in thread
From: Peter Maydell @ 2018-05-11 12:39 UTC (permalink / raw)
To: Paolo Bonzini; +Cc: QEMU Developers
On 11 May 2018 at 13:33, Paolo Bonzini <pbonzini@redhat.com> wrote:
> On 11/05/2018 14:19, Peter Maydell wrote:
>> Some of my build setups barf on the new glib version requirement:
>>
>> * my aarch64 build host (a gcc compile farm machine which is running
>> Ubuntu 14.04.5 LTS and has glib 2.40.2)
>> * my windows cross compile setups (which have glib 2.34.3)
>>
>> The latter could probably be fixed, but it will be a pain because
>> gtk.org don't distribute prebuilt glib &c zipfiles usable for cross
>> compile any more. For the former, that LTS is apparently still
>> in support for another year.
>
> Fair enough. I suppose that for now we can still uniformly bump up to
> 2.34, which is enough to remove all compatibility shims except for a couple.
>
> I'll resend the pull request without that part.
Thanks. If somebody has a recipe for doing mingw32/w64 cross builds
on Ubuntu hosts with a newer glib then I can look at updating
that bit of my build setup...
-- PMM
^ permalink raw reply [flat|nested] 40+ messages in thread
* Re: [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09
2018-05-11 12:19 ` [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09 Peter Maydell
2018-05-11 12:33 ` Paolo Bonzini
@ 2018-05-11 12:42 ` Daniel P. Berrangé
2018-05-11 12:50 ` Peter Maydell
1 sibling, 1 reply; 40+ messages in thread
From: Daniel P. Berrangé @ 2018-05-11 12:42 UTC (permalink / raw)
To: Peter Maydell; +Cc: Paolo Bonzini, QEMU Developers
On Fri, May 11, 2018 at 01:19:51PM +0100, Peter Maydell wrote:
> On 8 May 2018 at 23:14, Paolo Bonzini <pbonzini@redhat.com> wrote:
> > The following changes since commit cc8f8ba754bba17eea9791d67b572eb26e30b4ce:
> >
> > Merge remote-tracking branch 'remotes/ehabkost/tags/machine-next-pull-request' into staging (2018-05-08 15:25:17 +0100)
> >
> > are available in the git repository at:
> >
> >
> > git://github.com/bonzini/qemu.git tags/for-upstream
> >
> > for you to fetch changes up to 3c128df2ec239f5c1abd0aa59254c433a779a347:
> >
> > rename included C files to foo.inc.c, remove osdep.h (2018-05-09 00:13:41 +0200)
> >
> > ----------------------------------------------------------------
> > * Don't silently truncate extremely long words in the command line
> > * bump min required glib library version to 2.42
> > * dtc configure fixes
> > * MemoryRegionCache second try
> > * Deprecated option removal
> > * add support for Hyper-V reenlightenment MSRs
> >
> > ----------------------------------------------------------------
>
> Some of my build setups barf on the new glib version requirement:
>
> * my aarch64 build host (a gcc compile farm machine which is running
> Ubuntu 14.04.5 LTS and has glib 2.40.2)
That is expected - per the policy we only aim to support the most recent,
and most recent but one for 2 years overlap. Since LTS releases are two
years apart, this basically means always just 2 most recent LTS releases
for Ubuntu. Since we Ubuntu 18.04 LTS came out last month, we don't
need to continue 14.04 LTS
> * my windows cross compile setups (which have glib 2.34.3)
Presumably you're using a Debian host for that ?
The debian-win64-cross.docker job seems to pull in packages from
deb http://pkg.mxe.cc/repos/apt/debian jessie main
And mxe claims to have glib 2.54
Alternativel Fedora has new mingw packags too
> The latter could probably be fixed, but it will be a pain because
> gtk.org don't distribute prebuilt glib &c zipfiles usable for cross
> compile any more. For the former, that LTS is apparently still
> in support for another year.
Regards,
Daniel
--
|: https://berrange.com -o- https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org -o- https://fstop138.berrange.com :|
|: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|
^ permalink raw reply [flat|nested] 40+ messages in thread
* Re: [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09
2018-05-11 12:42 ` Daniel P. Berrangé
@ 2018-05-11 12:50 ` Peter Maydell
2018-05-11 12:54 ` Daniel P. Berrangé
0 siblings, 1 reply; 40+ messages in thread
From: Peter Maydell @ 2018-05-11 12:50 UTC (permalink / raw)
To: Daniel P. Berrangé; +Cc: Paolo Bonzini, QEMU Developers
On 11 May 2018 at 13:42, Daniel P. Berrangé <berrange@redhat.com> wrote:
> On Fri, May 11, 2018 at 01:19:51PM +0100, Peter Maydell wrote:
>> Some of my build setups barf on the new glib version requirement:
>>
>> * my aarch64 build host (a gcc compile farm machine which is running
>> Ubuntu 14.04.5 LTS and has glib 2.40.2)
>
> That is expected - per the policy we only aim to support the most recent,
> and most recent but one for 2 years overlap. Since LTS releases are two
> years apart, this basically means always just 2 most recent LTS releases
> for Ubuntu. Since we Ubuntu 18.04 LTS came out last month, we don't
> need to continue 14.04 LTS
...well, I could ask the GCC compile farm admins what their
upgrade plans are, but I don't particularly see the need
to impose load on a volunteer set of admins when the distro
they're running is still supported for another year. And I
don't want to drop aarch64 from my build test set.
>> * my windows cross compile setups (which have glib 2.34.3)
>
> Presumably you're using a Debian host for that ?
Ubuntu.
> The debian-win64-cross.docker job seems to pull in packages from
>
> deb http://pkg.mxe.cc/repos/apt/debian jessie main
>
> And mxe claims to have glib 2.54
Thanks, I will have a look at this. (They seem to have tarballs
as well as debian packages.)
-- PMM
^ permalink raw reply [flat|nested] 40+ messages in thread
* Re: [Qemu-devel] [PULL 00/30] Misc patches for 2018-05-09
2018-05-11 12:50 ` Peter Maydell
@ 2018-05-11 12:54 ` Daniel P. Berrangé
0 siblings, 0 replies; 40+ messages in thread
From: Daniel P. Berrangé @ 2018-05-11 12:54 UTC (permalink / raw)
To: Peter Maydell; +Cc: Paolo Bonzini, QEMU Developers
On Fri, May 11, 2018 at 01:50:28PM +0100, Peter Maydell wrote:
> On 11 May 2018 at 13:42, Daniel P. Berrangé <berrange@redhat.com> wrote:
> > On Fri, May 11, 2018 at 01:19:51PM +0100, Peter Maydell wrote:
> >> Some of my build setups barf on the new glib version requirement:
> >>
> >> * my aarch64 build host (a gcc compile farm machine which is running
> >> Ubuntu 14.04.5 LTS and has glib 2.40.2)
> >
> > That is expected - per the policy we only aim to support the most recent,
> > and most recent but one for 2 years overlap. Since LTS releases are two
> > years apart, this basically means always just 2 most recent LTS releases
> > for Ubuntu. Since we Ubuntu 18.04 LTS came out last month, we don't
> > need to continue 14.04 LTS
>
> ...well, I could ask the GCC compile farm admins what their
> upgrade plans are, but I don't particularly see the need
> to impose load on a volunteer set of admins when the distro
> they're running is still supported for another year. And I
> don't want to drop aarch64 from my build test set.
Ah ok, I didn't realize you weren't in control of the OS version.
2.40.2 is not too old, so we can accomodate that without significant
problems.
> >> * my windows cross compile setups (which have glib 2.34.3)
> >
> > Presumably you're using a Debian host for that ?
>
> Ubuntu.
>
> > The debian-win64-cross.docker job seems to pull in packages from
> >
> > deb http://pkg.mxe.cc/repos/apt/debian jessie main
> >
> > And mxe claims to have glib 2.54
>
> Thanks, I will have a look at this. (They seem to have tarballs
> as well as debian packages.)
Ok, let me know if that works out, and if so, I'll repost min glib
patches setting it to 2.40.2 to allow for the aarch64 build host.
Regards,
Daniel
--
|: https://berrange.com -o- https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org -o- https://fstop138.berrange.com :|
|: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|
^ permalink raw reply [flat|nested] 40+ messages in thread