[Qemu-devel] [PATCH v3 0/2] s390x/kvm: Properly detect page size of initial memory

[Qemu-devel] [PATCH v3 0/2] s390x/kvm: Properly detect page size of initial memory

[David Hildenbrand]

While the detection on s390x of the maximum ram pagesize works by pure luck,
we should try to do better. Especially, once "-mem-path" is replaced
by memory backends, the current approach will break.

Configure the maximum page size after configuring initital memory. Provide
a function to detect the maximum ram pagesize

I think this can be deferred to 4.1 as I don't think anything is actually
broken right now.

v2 -> v3:
- Fix compile error
- Tested that hugepages correctly result in an error with
  "modprobe kvm hpage=0", but not with "modprobe kvm hpage=1".

v1 -> v2:
- Add a better comment describing that only initial memory is considered
- Add "exec: Introduce qemu_getmaxrampagesize() and rename
  qemu_getrampagesize()"

David Hildenbrand (2):
  s390x/kvm: Configure page size after memory has actually been
    initialized
  exec: Introduce qemu_maxrampagesize() and rename qemu_getrampagesize()

 exec.c                     | 44 ++++++++++++++++++++++++++++++++++----
 hw/ppc/spapr_caps.c        |  4 ++--
 hw/s390x/s390-virtio-ccw.c | 12 +++++++++++
 hw/vfio/spapr.c            |  2 +-
 include/exec/ram_addr.h    |  3 ++-
 target/ppc/kvm.c           |  2 +-
 target/s390x/cpu.c         |  7 ++++++
 target/s390x/cpu.h         |  1 +
 target/s390x/kvm-stub.c    |  4 ++++
 target/s390x/kvm.c         | 35 ++++++++++++------------------
 target/s390x/kvm_s390x.h   |  1 +
 11 files changed, 85 insertions(+), 30 deletions(-)

-- 
2.20.1

[Qemu-devel] [PATCH v3 0/2] s390x/kvm: Properly detect page size of initial memory

[David Hildenbrand]

While the detection on s390x of the maximum ram pagesize works by pure luck,
we should try to do better. Especially, once "-mem-path" is replaced
by memory backends, the current approach will break.

Configure the maximum page size after configuring initital memory. Provide
a function to detect the maximum ram pagesize

I think this can be deferred to 4.1 as I don't think anything is actually
broken right now.

v2 -> v3:
- Fix compile error
- Tested that hugepages correctly result in an error with
  "modprobe kvm hpage=0", but not with "modprobe kvm hpage=1".

v1 -> v2:
- Add a better comment describing that only initial memory is considered
- Add "exec: Introduce qemu_getmaxrampagesize() and rename
  qemu_getrampagesize()"

David Hildenbrand (2):
  s390x/kvm: Configure page size after memory has actually been
    initialized
  exec: Introduce qemu_maxrampagesize() and rename qemu_getrampagesize()

 exec.c                     | 44 ++++++++++++++++++++++++++++++++++----
 hw/ppc/spapr_caps.c        |  4 ++--
 hw/s390x/s390-virtio-ccw.c | 12 +++++++++++
 hw/vfio/spapr.c            |  2 +-
 include/exec/ram_addr.h    |  3 ++-
 target/ppc/kvm.c           |  2 +-
 target/s390x/cpu.c         |  7 ++++++
 target/s390x/cpu.h         |  1 +
 target/s390x/kvm-stub.c    |  4 ++++
 target/s390x/kvm.c         | 35 ++++++++++++------------------
 target/s390x/kvm_s390x.h   |  1 +
 11 files changed, 85 insertions(+), 30 deletions(-)

-- 
2.20.1

[Qemu-devel] [PATCH v3 1/2] s390x/kvm: Configure page size after memory has actually been initialized

[David Hildenbrand]

Right now we configure the pagesize quite early, when initializing KVM.
This is long before system memory is actually allocated via
memory_region_allocate_system_memory(), and therefore memory backends
marked as mapped.

Instead, let's configure the maximum page size after initializing
memory in s390_memory_init(). cap_hpage_1m is still properly
configured before creating any CPUs, and therefore before configuring
the CPU model and eventually enabling CMMA.

This is not a fix but rather a preparation for the future, when initial
memory might reside on memory backends (not the case for s390x right now)
We will replace qemu_getrampagesize() soon by a function that will always
return the maximum page size (not the minimum page size, which only
works by pure luck so far, as there are no memory backends).

Acked-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: David Hildenbrand <david@redhat.com>
---
 hw/s390x/s390-virtio-ccw.c | 12 ++++++++++++
 target/s390x/cpu.c         |  7 +++++++
 target/s390x/cpu.h         |  1 +
 target/s390x/kvm-stub.c    |  4 ++++
 target/s390x/kvm.c         | 35 ++++++++++++++---------------------
 target/s390x/kvm_s390x.h   |  1 +
 6 files changed, 39 insertions(+), 21 deletions(-)

diff --git a/hw/s390x/s390-virtio-ccw.c b/hw/s390x/s390-virtio-ccw.c
index d11069b860..3be5679657 100644
--- a/hw/s390x/s390-virtio-ccw.c
+++ b/hw/s390x/s390-virtio-ccw.c
@@ -15,6 +15,7 @@
 #include "cpu.h"
 #include "hw/boards.h"
 #include "exec/address-spaces.h"
+#include "exec/ram_addr.h"
 #include "hw/s390x/s390-virtio-hcall.h"
 #include "hw/s390x/sclp.h"
 #include "hw/s390x/s390_flic.h"
@@ -163,6 +164,7 @@ static void s390_memory_init(ram_addr_t mem_size)
     MemoryRegion *sysmem = get_system_memory();
     ram_addr_t chunk, offset = 0;
     unsigned int number = 0;
+    Error *local_err = NULL;
     gchar *name;
 
     /* allocate RAM for core */
@@ -182,6 +184,15 @@ static void s390_memory_init(ram_addr_t mem_size)
     }
     g_free(name);
 
+    /*
+     * Configure the maximum page size. As no memory devices were created
+     * yet, this is the page size of initial memory only.
+     */
+    s390_set_max_pagesize(qemu_getrampagesize(), &local_err);
+    if (local_err) {
+        error_report_err(local_err);
+        exit(EXIT_FAILURE);
+    }
     /* Initialize storage key device */
     s390_skeys_init();
     /* Initialize storage attributes device */
@@ -253,6 +264,7 @@ static void ccw_init(MachineState *machine)
     DeviceState *dev;
 
     s390_sclp_init();
+    /* init memory + setup max page size. Required for the CPU model */
     s390_memory_init(machine->ram_size);
 
     /* init CPUs (incl. CPU model) early so s390_has_feature() works */
diff --git a/target/s390x/cpu.c b/target/s390x/cpu.c
index 698dd9cb82..b58ef0a8ef 100644
--- a/target/s390x/cpu.c
+++ b/target/s390x/cpu.c
@@ -399,6 +399,13 @@ int s390_set_memory_limit(uint64_t new_limit, uint64_t *hw_limit)
     return 0;
 }
 
+void s390_set_max_pagesize(uint64_t pagesize, Error **errp)
+{
+    if (kvm_enabled()) {
+        kvm_s390_set_max_pagesize(pagesize, errp);
+    }
+}
+
 void s390_cmma_reset(void)
 {
     if (kvm_enabled()) {
diff --git a/target/s390x/cpu.h b/target/s390x/cpu.h
index cb6d77053a..c14be2b5ba 100644
--- a/target/s390x/cpu.h
+++ b/target/s390x/cpu.h
@@ -734,6 +734,7 @@ static inline void s390_do_cpu_load_normal(CPUState *cs, run_on_cpu_data arg)
 /* cpu.c */
 void s390_crypto_reset(void);
 int s390_set_memory_limit(uint64_t new_limit, uint64_t *hw_limit);
+void s390_set_max_pagesize(uint64_t pagesize, Error **errp);
 void s390_cmma_reset(void);
 void s390_enable_css_support(S390CPU *cpu);
 int s390_assign_subch_ioeventfd(EventNotifier *notifier, uint32_t sch_id,
diff --git a/target/s390x/kvm-stub.c b/target/s390x/kvm-stub.c
index bf7795e47a..22b4514ca6 100644
--- a/target/s390x/kvm-stub.c
+++ b/target/s390x/kvm-stub.c
@@ -93,6 +93,10 @@ int kvm_s390_set_mem_limit(uint64_t new_limit, uint64_t *hw_limit)
     return 0;
 }
 
+void kvm_s390_set_max_pagesize(uint64_t pagesize, Error **errp)
+{
+}
+
 void kvm_s390_crypto_reset(void)
 {
 }
diff --git a/target/s390x/kvm.c b/target/s390x/kvm.c
index 19530fb94e..bee73dc1a4 100644
--- a/target/s390x/kvm.c
+++ b/target/s390x/kvm.c
@@ -283,44 +283,37 @@ void kvm_s390_crypto_reset(void)
     }
 }
 
-static int kvm_s390_configure_mempath_backing(KVMState *s)
+void kvm_s390_set_max_pagesize(uint64_t pagesize, Error **errp)
 {
-    size_t path_psize = qemu_getrampagesize();
-
-    if (path_psize == 4 * KiB) {
-        return 0;
+    if (pagesize == 4 * KiB) {
+        return;
     }
 
     if (!hpage_1m_allowed()) {
-        error_report("This QEMU machine does not support huge page "
-                     "mappings");
-        return -EINVAL;
+        error_setg(errp, "This QEMU machine does not support huge page "
+                   "mappings");
+        return;
     }
 
-    if (path_psize != 1 * MiB) {
-        error_report("Memory backing with 2G pages was specified, "
-                     "but KVM does not support this memory backing");
-        return -EINVAL;
+    if (pagesize != 1 * MiB) {
+        error_setg(errp, "Memory backing with 2G pages was specified, "
+                   "but KVM does not support this memory backing");
+        return;
     }
 
-    if (kvm_vm_enable_cap(s, KVM_CAP_S390_HPAGE_1M, 0)) {
-        error_report("Memory backing with 1M pages was specified, "
-                     "but KVM does not support this memory backing");
-        return -EINVAL;
+    if (kvm_vm_enable_cap(kvm_state, KVM_CAP_S390_HPAGE_1M, 0)) {
+        error_setg(errp, "Memory backing with 1M pages was specified, "
+                   "but KVM does not support this memory backing");
+        return;
     }
 
     cap_hpage_1m = 1;
-    return 0;
 }
 
 int kvm_arch_init(MachineState *ms, KVMState *s)
 {
     MachineClass *mc = MACHINE_GET_CLASS(ms);
 
-    if (kvm_s390_configure_mempath_backing(s)) {
-        return -EINVAL;
-    }
-
     mc->default_cpu_type = S390_CPU_TYPE_NAME("host");
     cap_sync_regs = kvm_check_extension(s, KVM_CAP_SYNC_REGS);
     cap_async_pf = kvm_check_extension(s, KVM_CAP_ASYNC_PF);
diff --git a/target/s390x/kvm_s390x.h b/target/s390x/kvm_s390x.h
index 6e52287da3..caf985955b 100644
--- a/target/s390x/kvm_s390x.h
+++ b/target/s390x/kvm_s390x.h
@@ -36,6 +36,7 @@ int kvm_s390_cmma_active(void);
 void kvm_s390_cmma_reset(void);
 void kvm_s390_reset_vcpu(S390CPU *cpu);
 int kvm_s390_set_mem_limit(uint64_t new_limit, uint64_t *hw_limit);
+void kvm_s390_set_max_pagesize(uint64_t pagesize, Error **errp);
 void kvm_s390_crypto_reset(void);
 void kvm_s390_restart_interrupt(S390CPU *cpu);
 void kvm_s390_stop_interrupt(S390CPU *cpu);
-- 
2.20.1

[Qemu-devel] [PATCH v3 1/2] s390x/kvm: Configure page size after memory has actually been initialized

[David Hildenbrand]

Right now we configure the pagesize quite early, when initializing KVM.
This is long before system memory is actually allocated via
memory_region_allocate_system_memory(), and therefore memory backends
marked as mapped.

Instead, let's configure the maximum page size after initializing
memory in s390_memory_init(). cap_hpage_1m is still properly
configured before creating any CPUs, and therefore before configuring
the CPU model and eventually enabling CMMA.

This is not a fix but rather a preparation for the future, when initial
memory might reside on memory backends (not the case for s390x right now)
We will replace qemu_getrampagesize() soon by a function that will always
return the maximum page size (not the minimum page size, which only
works by pure luck so far, as there are no memory backends).

Acked-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: David Hildenbrand <david@redhat.com>
---
 hw/s390x/s390-virtio-ccw.c | 12 ++++++++++++
 target/s390x/cpu.c         |  7 +++++++
 target/s390x/cpu.h         |  1 +
 target/s390x/kvm-stub.c    |  4 ++++
 target/s390x/kvm.c         | 35 ++++++++++++++---------------------
 target/s390x/kvm_s390x.h   |  1 +
 6 files changed, 39 insertions(+), 21 deletions(-)

diff --git a/hw/s390x/s390-virtio-ccw.c b/hw/s390x/s390-virtio-ccw.c
index d11069b860..3be5679657 100644
--- a/hw/s390x/s390-virtio-ccw.c
+++ b/hw/s390x/s390-virtio-ccw.c
@@ -15,6 +15,7 @@
 #include "cpu.h"
 #include "hw/boards.h"
 #include "exec/address-spaces.h"
+#include "exec/ram_addr.h"
 #include "hw/s390x/s390-virtio-hcall.h"
 #include "hw/s390x/sclp.h"
 #include "hw/s390x/s390_flic.h"
@@ -163,6 +164,7 @@ static void s390_memory_init(ram_addr_t mem_size)
     MemoryRegion *sysmem = get_system_memory();
     ram_addr_t chunk, offset = 0;
     unsigned int number = 0;
+    Error *local_err = NULL;
     gchar *name;
 
     /* allocate RAM for core */
@@ -182,6 +184,15 @@ static void s390_memory_init(ram_addr_t mem_size)
     }
     g_free(name);
 
+    /*
+     * Configure the maximum page size. As no memory devices were created
+     * yet, this is the page size of initial memory only.
+     */
+    s390_set_max_pagesize(qemu_getrampagesize(), &local_err);
+    if (local_err) {
+        error_report_err(local_err);
+        exit(EXIT_FAILURE);
+    }
     /* Initialize storage key device */
     s390_skeys_init();
     /* Initialize storage attributes device */
@@ -253,6 +264,7 @@ static void ccw_init(MachineState *machine)
     DeviceState *dev;
 
     s390_sclp_init();
+    /* init memory + setup max page size. Required for the CPU model */
     s390_memory_init(machine->ram_size);
 
     /* init CPUs (incl. CPU model) early so s390_has_feature() works */
diff --git a/target/s390x/cpu.c b/target/s390x/cpu.c
index 698dd9cb82..b58ef0a8ef 100644
--- a/target/s390x/cpu.c
+++ b/target/s390x/cpu.c
@@ -399,6 +399,13 @@ int s390_set_memory_limit(uint64_t new_limit, uint64_t *hw_limit)
     return 0;
 }
 
+void s390_set_max_pagesize(uint64_t pagesize, Error **errp)
+{
+    if (kvm_enabled()) {
+        kvm_s390_set_max_pagesize(pagesize, errp);
+    }
+}
+
 void s390_cmma_reset(void)
 {
     if (kvm_enabled()) {
diff --git a/target/s390x/cpu.h b/target/s390x/cpu.h
index cb6d77053a..c14be2b5ba 100644
--- a/target/s390x/cpu.h
+++ b/target/s390x/cpu.h
@@ -734,6 +734,7 @@ static inline void s390_do_cpu_load_normal(CPUState *cs, run_on_cpu_data arg)
 /* cpu.c */
 void s390_crypto_reset(void);
 int s390_set_memory_limit(uint64_t new_limit, uint64_t *hw_limit);
+void s390_set_max_pagesize(uint64_t pagesize, Error **errp);
 void s390_cmma_reset(void);
 void s390_enable_css_support(S390CPU *cpu);
 int s390_assign_subch_ioeventfd(EventNotifier *notifier, uint32_t sch_id,
diff --git a/target/s390x/kvm-stub.c b/target/s390x/kvm-stub.c
index bf7795e47a..22b4514ca6 100644
--- a/target/s390x/kvm-stub.c
+++ b/target/s390x/kvm-stub.c
@@ -93,6 +93,10 @@ int kvm_s390_set_mem_limit(uint64_t new_limit, uint64_t *hw_limit)
     return 0;
 }
 
+void kvm_s390_set_max_pagesize(uint64_t pagesize, Error **errp)
+{
+}
+
 void kvm_s390_crypto_reset(void)
 {
 }
diff --git a/target/s390x/kvm.c b/target/s390x/kvm.c
index 19530fb94e..bee73dc1a4 100644
--- a/target/s390x/kvm.c
+++ b/target/s390x/kvm.c
@@ -283,44 +283,37 @@ void kvm_s390_crypto_reset(void)
     }
 }
 
-static int kvm_s390_configure_mempath_backing(KVMState *s)
+void kvm_s390_set_max_pagesize(uint64_t pagesize, Error **errp)
 {
-    size_t path_psize = qemu_getrampagesize();
-
-    if (path_psize == 4 * KiB) {
-        return 0;
+    if (pagesize == 4 * KiB) {
+        return;
     }
 
     if (!hpage_1m_allowed()) {
-        error_report("This QEMU machine does not support huge page "
-                     "mappings");
-        return -EINVAL;
+        error_setg(errp, "This QEMU machine does not support huge page "
+                   "mappings");
+        return;
     }
 
-    if (path_psize != 1 * MiB) {
-        error_report("Memory backing with 2G pages was specified, "
-                     "but KVM does not support this memory backing");
-        return -EINVAL;
+    if (pagesize != 1 * MiB) {
+        error_setg(errp, "Memory backing with 2G pages was specified, "
+                   "but KVM does not support this memory backing");
+        return;
     }
 
-    if (kvm_vm_enable_cap(s, KVM_CAP_S390_HPAGE_1M, 0)) {
-        error_report("Memory backing with 1M pages was specified, "
-                     "but KVM does not support this memory backing");
-        return -EINVAL;
+    if (kvm_vm_enable_cap(kvm_state, KVM_CAP_S390_HPAGE_1M, 0)) {
+        error_setg(errp, "Memory backing with 1M pages was specified, "
+                   "but KVM does not support this memory backing");
+        return;
     }
 
     cap_hpage_1m = 1;
-    return 0;
 }
 
 int kvm_arch_init(MachineState *ms, KVMState *s)
 {
     MachineClass *mc = MACHINE_GET_CLASS(ms);
 
-    if (kvm_s390_configure_mempath_backing(s)) {
-        return -EINVAL;
-    }
-
     mc->default_cpu_type = S390_CPU_TYPE_NAME("host");
     cap_sync_regs = kvm_check_extension(s, KVM_CAP_SYNC_REGS);
     cap_async_pf = kvm_check_extension(s, KVM_CAP_ASYNC_PF);
diff --git a/target/s390x/kvm_s390x.h b/target/s390x/kvm_s390x.h
index 6e52287da3..caf985955b 100644
--- a/target/s390x/kvm_s390x.h
+++ b/target/s390x/kvm_s390x.h
@@ -36,6 +36,7 @@ int kvm_s390_cmma_active(void);
 void kvm_s390_cmma_reset(void);
 void kvm_s390_reset_vcpu(S390CPU *cpu);
 int kvm_s390_set_mem_limit(uint64_t new_limit, uint64_t *hw_limit);
+void kvm_s390_set_max_pagesize(uint64_t pagesize, Error **errp);
 void kvm_s390_crypto_reset(void);
 void kvm_s390_restart_interrupt(S390CPU *cpu);
 void kvm_s390_stop_interrupt(S390CPU *cpu);
-- 
2.20.1

[Qemu-devel] [PATCH v3 2/2] exec: Introduce qemu_maxrampagesize() and rename qemu_getrampagesize()

[David Hildenbrand]

Rename qemu_getrampagesize() to qemu_minrampagesize(). While at it,
properly rename find_max_supported_pagesize() to
find_min_backend_pagesize().

s390x is actually interested into the maximum ram pagesize, so
introduce and use qemu_maxrampagesize().

Add a TODO, indicating that looking at any mapped memory backends is not
100% correct in some cases.

Signed-off-by: David Hildenbrand <david@redhat.com>
---
 exec.c                     | 44 ++++++++++++++++++++++++++++++++++----
 hw/ppc/spapr_caps.c        |  4 ++--
 hw/s390x/s390-virtio-ccw.c |  2 +-
 hw/vfio/spapr.c            |  2 +-
 include/exec/ram_addr.h    |  3 ++-
 target/ppc/kvm.c           |  2 +-
 6 files changed, 47 insertions(+), 10 deletions(-)

diff --git a/exec.c b/exec.c
index 6ab62f4eee..f9b834ef82 100644
--- a/exec.c
+++ b/exec.c
@@ -1687,7 +1687,7 @@ void ram_block_dump(Monitor *mon)
  * when we actually open and map them.  Iterate over the file
  * descriptors instead, and use qemu_fd_getpagesize().
  */
-static int find_max_supported_pagesize(Object *obj, void *opaque)
+static int find_min_backend_pagesize(Object *obj, void *opaque)
 {
     long *hpsize_min = opaque;
 
@@ -1703,7 +1703,27 @@ static int find_max_supported_pagesize(Object *obj, void *opaque)
     return 0;
 }
 
-long qemu_getrampagesize(void)
+static int find_max_backend_pagesize(Object *obj, void *opaque)
+{
+    long *hpsize_max = opaque;
+
+    if (object_dynamic_cast(obj, TYPE_MEMORY_BACKEND)) {
+        HostMemoryBackend *backend = MEMORY_BACKEND(obj);
+        long hpsize = host_memory_backend_pagesize(backend);
+
+        if (host_memory_backend_is_mapped(backend) && (hpsize > *hpsize_max)) {
+            *hpsize_max = hpsize;
+        }
+    }
+
+    return 0;
+}
+
+/*
+ * TODO: We assume right now that all mapped host memory backends are
+ * used as RAM, however some might be used for different purposes.
+ */
+long qemu_minrampagesize(void)
 {
     long hpsize = LONG_MAX;
     long mainrampagesize;
@@ -1723,7 +1743,7 @@ long qemu_getrampagesize(void)
      */
     memdev_root = object_resolve_path("/objects", NULL);
     if (memdev_root) {
-        object_child_foreach(memdev_root, find_max_supported_pagesize, &hpsize);
+        object_child_foreach(memdev_root, find_min_backend_pagesize, &hpsize);
     }
     if (hpsize == LONG_MAX) {
         /* No additional memory regions found ==> Report main RAM page size */
@@ -1746,8 +1766,24 @@ long qemu_getrampagesize(void)
 
     return hpsize;
 }
+
+long qemu_maxrampagesize(void)
+{
+    long pagesize = qemu_mempath_getpagesize(mem_path);
+    Object *memdev_root = object_resolve_path("/objects", NULL);
+
+    if (memdev_root) {
+        object_child_foreach(memdev_root, find_max_backend_pagesize,
+                             &pagesize);
+    }
+    return pagesize;
+}
 #else
-long qemu_getrampagesize(void)
+long qemu_minrampagesize(void)
+{
+    return getpagesize();
+}
+long qemu_maxrampagesize(void)
 {
     return getpagesize();
 }
diff --git a/hw/ppc/spapr_caps.c b/hw/ppc/spapr_caps.c
index edc5ed0e0c..9b1c10baa6 100644
--- a/hw/ppc/spapr_caps.c
+++ b/hw/ppc/spapr_caps.c
@@ -347,7 +347,7 @@ static void cap_hpt_maxpagesize_apply(SpaprMachineState *spapr,
         warn_report("Many guests require at least 64kiB hpt-max-page-size");
     }
 
-    spapr_check_pagesize(spapr, qemu_getrampagesize(), errp);
+    spapr_check_pagesize(spapr, qemu_minrampagesize(), errp);
 }
 
 static bool spapr_pagesize_cb(void *opaque, uint32_t seg_pshift,
@@ -609,7 +609,7 @@ static SpaprCapabilities default_caps_with_cpu(SpaprMachineState *spapr,
         uint8_t mps;
 
         if (kvmppc_hpt_needs_host_contiguous_pages()) {
-            mps = ctz64(qemu_getrampagesize());
+            mps = ctz64(qemu_minrampagesize());
         } else {
             mps = 34; /* allow everything up to 16GiB, i.e. everything */
         }
diff --git a/hw/s390x/s390-virtio-ccw.c b/hw/s390x/s390-virtio-ccw.c
index 3be5679657..7e256d3d31 100644
--- a/hw/s390x/s390-virtio-ccw.c
+++ b/hw/s390x/s390-virtio-ccw.c
@@ -188,7 +188,7 @@ static void s390_memory_init(ram_addr_t mem_size)
      * Configure the maximum page size. As no memory devices were created
      * yet, this is the page size of initial memory only.
      */
-    s390_set_max_pagesize(qemu_getrampagesize(), &local_err);
+    s390_set_max_pagesize(qemu_maxrampagesize(), &local_err);
     if (local_err) {
         error_report_err(local_err);
         exit(EXIT_FAILURE);
diff --git a/hw/vfio/spapr.c b/hw/vfio/spapr.c
index 57fe758e54..96c0ad9d9b 100644
--- a/hw/vfio/spapr.c
+++ b/hw/vfio/spapr.c
@@ -148,7 +148,7 @@ int vfio_spapr_create_window(VFIOContainer *container,
     uint64_t pagesize = memory_region_iommu_get_min_page_size(iommu_mr);
     unsigned entries, bits_total, bits_per_level, max_levels;
     struct vfio_iommu_spapr_tce_create create = { .argsz = sizeof(create) };
-    long rampagesize = qemu_getrampagesize();
+    long rampagesize = qemu_minrampagesize();
 
     /*
      * The host might not support the guest supported IOMMU page size,
diff --git a/include/exec/ram_addr.h b/include/exec/ram_addr.h
index 9ecd911c3e..139ad79390 100644
--- a/include/exec/ram_addr.h
+++ b/include/exec/ram_addr.h
@@ -73,7 +73,8 @@ static inline unsigned long int ramblock_recv_bitmap_offset(void *host_addr,
 
 bool ramblock_is_pmem(RAMBlock *rb);
 
-long qemu_getrampagesize(void);
+long qemu_minrampagesize(void);
+long qemu_maxrampagesize(void);
 
 /**
  * qemu_ram_alloc_from_file,
diff --git a/target/ppc/kvm.c b/target/ppc/kvm.c
index 2427c8ee13..59d92c4275 100644
--- a/target/ppc/kvm.c
+++ b/target/ppc/kvm.c
@@ -2136,7 +2136,7 @@ uint64_t kvmppc_rma_size(uint64_t current_size, unsigned int hash_shift)
     /* Find the largest hardware supported page size that's less than
      * or equal to the (logical) backing page size of guest RAM */
     kvm_get_smmu_info(&info, &error_fatal);
-    rampagesize = qemu_getrampagesize();
+    rampagesize = qemu_minrampagesize();
     best_page_shift = 0;
 
     for (i = 0; i < KVM_PPC_PAGE_SIZES_MAX_SZ; i++) {
-- 
2.20.1

[Qemu-devel] [PATCH v3 2/2] exec: Introduce qemu_maxrampagesize() and rename qemu_getrampagesize()

[David Hildenbrand]

Rename qemu_getrampagesize() to qemu_minrampagesize(). While at it,
properly rename find_max_supported_pagesize() to
find_min_backend_pagesize().

s390x is actually interested into the maximum ram pagesize, so
introduce and use qemu_maxrampagesize().

Add a TODO, indicating that looking at any mapped memory backends is not
100% correct in some cases.

Signed-off-by: David Hildenbrand <david@redhat.com>
---
 exec.c                     | 44 ++++++++++++++++++++++++++++++++++----
 hw/ppc/spapr_caps.c        |  4 ++--
 hw/s390x/s390-virtio-ccw.c |  2 +-
 hw/vfio/spapr.c            |  2 +-
 include/exec/ram_addr.h    |  3 ++-
 target/ppc/kvm.c           |  2 +-
 6 files changed, 47 insertions(+), 10 deletions(-)

diff --git a/exec.c b/exec.c
index 6ab62f4eee..f9b834ef82 100644
--- a/exec.c
+++ b/exec.c
@@ -1687,7 +1687,7 @@ void ram_block_dump(Monitor *mon)
  * when we actually open and map them.  Iterate over the file
  * descriptors instead, and use qemu_fd_getpagesize().
  */
-static int find_max_supported_pagesize(Object *obj, void *opaque)
+static int find_min_backend_pagesize(Object *obj, void *opaque)
 {
     long *hpsize_min = opaque;
 
@@ -1703,7 +1703,27 @@ static int find_max_supported_pagesize(Object *obj, void *opaque)
     return 0;
 }
 
-long qemu_getrampagesize(void)
+static int find_max_backend_pagesize(Object *obj, void *opaque)
+{
+    long *hpsize_max = opaque;
+
+    if (object_dynamic_cast(obj, TYPE_MEMORY_BACKEND)) {
+        HostMemoryBackend *backend = MEMORY_BACKEND(obj);
+        long hpsize = host_memory_backend_pagesize(backend);
+
+        if (host_memory_backend_is_mapped(backend) && (hpsize > *hpsize_max)) {
+            *hpsize_max = hpsize;
+        }
+    }
+
+    return 0;
+}
+
+/*
+ * TODO: We assume right now that all mapped host memory backends are
+ * used as RAM, however some might be used for different purposes.
+ */
+long qemu_minrampagesize(void)
 {
     long hpsize = LONG_MAX;
     long mainrampagesize;
@@ -1723,7 +1743,7 @@ long qemu_getrampagesize(void)
      */
     memdev_root = object_resolve_path("/objects", NULL);
     if (memdev_root) {
-        object_child_foreach(memdev_root, find_max_supported_pagesize, &hpsize);
+        object_child_foreach(memdev_root, find_min_backend_pagesize, &hpsize);
     }
     if (hpsize == LONG_MAX) {
         /* No additional memory regions found ==> Report main RAM page size */
@@ -1746,8 +1766,24 @@ long qemu_getrampagesize(void)
 
     return hpsize;
 }
+
+long qemu_maxrampagesize(void)
+{
+    long pagesize = qemu_mempath_getpagesize(mem_path);
+    Object *memdev_root = object_resolve_path("/objects", NULL);
+
+    if (memdev_root) {
+        object_child_foreach(memdev_root, find_max_backend_pagesize,
+                             &pagesize);
+    }
+    return pagesize;
+}
 #else
-long qemu_getrampagesize(void)
+long qemu_minrampagesize(void)
+{
+    return getpagesize();
+}
+long qemu_maxrampagesize(void)
 {
     return getpagesize();
 }
diff --git a/hw/ppc/spapr_caps.c b/hw/ppc/spapr_caps.c
index edc5ed0e0c..9b1c10baa6 100644
--- a/hw/ppc/spapr_caps.c
+++ b/hw/ppc/spapr_caps.c
@@ -347,7 +347,7 @@ static void cap_hpt_maxpagesize_apply(SpaprMachineState *spapr,
         warn_report("Many guests require at least 64kiB hpt-max-page-size");
     }
 
-    spapr_check_pagesize(spapr, qemu_getrampagesize(), errp);
+    spapr_check_pagesize(spapr, qemu_minrampagesize(), errp);
 }
 
 static bool spapr_pagesize_cb(void *opaque, uint32_t seg_pshift,
@@ -609,7 +609,7 @@ static SpaprCapabilities default_caps_with_cpu(SpaprMachineState *spapr,
         uint8_t mps;
 
         if (kvmppc_hpt_needs_host_contiguous_pages()) {
-            mps = ctz64(qemu_getrampagesize());
+            mps = ctz64(qemu_minrampagesize());
         } else {
             mps = 34; /* allow everything up to 16GiB, i.e. everything */
         }
diff --git a/hw/s390x/s390-virtio-ccw.c b/hw/s390x/s390-virtio-ccw.c
index 3be5679657..7e256d3d31 100644
--- a/hw/s390x/s390-virtio-ccw.c
+++ b/hw/s390x/s390-virtio-ccw.c
@@ -188,7 +188,7 @@ static void s390_memory_init(ram_addr_t mem_size)
      * Configure the maximum page size. As no memory devices were created
      * yet, this is the page size of initial memory only.
      */
-    s390_set_max_pagesize(qemu_getrampagesize(), &local_err);
+    s390_set_max_pagesize(qemu_maxrampagesize(), &local_err);
     if (local_err) {
         error_report_err(local_err);
         exit(EXIT_FAILURE);
diff --git a/hw/vfio/spapr.c b/hw/vfio/spapr.c
index 57fe758e54..96c0ad9d9b 100644
--- a/hw/vfio/spapr.c
+++ b/hw/vfio/spapr.c
@@ -148,7 +148,7 @@ int vfio_spapr_create_window(VFIOContainer *container,
     uint64_t pagesize = memory_region_iommu_get_min_page_size(iommu_mr);
     unsigned entries, bits_total, bits_per_level, max_levels;
     struct vfio_iommu_spapr_tce_create create = { .argsz = sizeof(create) };
-    long rampagesize = qemu_getrampagesize();
+    long rampagesize = qemu_minrampagesize();
 
     /*
      * The host might not support the guest supported IOMMU page size,
diff --git a/include/exec/ram_addr.h b/include/exec/ram_addr.h
index 9ecd911c3e..139ad79390 100644
--- a/include/exec/ram_addr.h
+++ b/include/exec/ram_addr.h
@@ -73,7 +73,8 @@ static inline unsigned long int ramblock_recv_bitmap_offset(void *host_addr,
 
 bool ramblock_is_pmem(RAMBlock *rb);
 
-long qemu_getrampagesize(void);
+long qemu_minrampagesize(void);
+long qemu_maxrampagesize(void);
 
 /**
  * qemu_ram_alloc_from_file,
diff --git a/target/ppc/kvm.c b/target/ppc/kvm.c
index 2427c8ee13..59d92c4275 100644
--- a/target/ppc/kvm.c
+++ b/target/ppc/kvm.c
@@ -2136,7 +2136,7 @@ uint64_t kvmppc_rma_size(uint64_t current_size, unsigned int hash_shift)
     /* Find the largest hardware supported page size that's less than
      * or equal to the (logical) backing page size of guest RAM */
     kvm_get_smmu_info(&info, &error_fatal);
-    rampagesize = qemu_getrampagesize();
+    rampagesize = qemu_minrampagesize();
     best_page_shift = 0;
 
     for (i = 0; i < KVM_PPC_PAGE_SIZES_MAX_SZ; i++) {
-- 
2.20.1

Re: [Qemu-devel] [PATCH v3 2/2] exec: Introduce qemu_maxrampagesize() and rename qemu_getrampagesize()

[David Gibson]

[-- Attachment #1: Type: text/plain, Size: 6908 bytes --]

On Wed, Apr 17, 2019 at 01:31:43PM +0200, David Hildenbrand wrote:
> Rename qemu_getrampagesize() to qemu_minrampagesize(). While at it,
> properly rename find_max_supported_pagesize() to
> find_min_backend_pagesize().
> 
> s390x is actually interested into the maximum ram pagesize, so
> introduce and use qemu_maxrampagesize().
> 
> Add a TODO, indicating that looking at any mapped memory backends is not
> 100% correct in some cases.
> 
> Signed-off-by: David Hildenbrand <david@redhat.com>

Reviewed-by: David Gibson <david@gibson.dropbear.id.au>

> ---
>  exec.c                     | 44 ++++++++++++++++++++++++++++++++++----
>  hw/ppc/spapr_caps.c        |  4 ++--
>  hw/s390x/s390-virtio-ccw.c |  2 +-
>  hw/vfio/spapr.c            |  2 +-
>  include/exec/ram_addr.h    |  3 ++-
>  target/ppc/kvm.c           |  2 +-
>  6 files changed, 47 insertions(+), 10 deletions(-)
> 
> diff --git a/exec.c b/exec.c
> index 6ab62f4eee..f9b834ef82 100644
> --- a/exec.c
> +++ b/exec.c
> @@ -1687,7 +1687,7 @@ void ram_block_dump(Monitor *mon)
>   * when we actually open and map them.  Iterate over the file
>   * descriptors instead, and use qemu_fd_getpagesize().
>   */
> -static int find_max_supported_pagesize(Object *obj, void *opaque)
> +static int find_min_backend_pagesize(Object *obj, void *opaque)
>  {
>      long *hpsize_min = opaque;
>  
> @@ -1703,7 +1703,27 @@ static int find_max_supported_pagesize(Object *obj, void *opaque)
>      return 0;
>  }
>  
> -long qemu_getrampagesize(void)
> +static int find_max_backend_pagesize(Object *obj, void *opaque)
> +{
> +    long *hpsize_max = opaque;
> +
> +    if (object_dynamic_cast(obj, TYPE_MEMORY_BACKEND)) {
> +        HostMemoryBackend *backend = MEMORY_BACKEND(obj);
> +        long hpsize = host_memory_backend_pagesize(backend);
> +
> +        if (host_memory_backend_is_mapped(backend) && (hpsize > *hpsize_max)) {
> +            *hpsize_max = hpsize;
> +        }
> +    }
> +
> +    return 0;
> +}
> +
> +/*
> + * TODO: We assume right now that all mapped host memory backends are
> + * used as RAM, however some might be used for different purposes.
> + */
> +long qemu_minrampagesize(void)
>  {
>      long hpsize = LONG_MAX;
>      long mainrampagesize;
> @@ -1723,7 +1743,7 @@ long qemu_getrampagesize(void)
>       */
>      memdev_root = object_resolve_path("/objects", NULL);
>      if (memdev_root) {
> -        object_child_foreach(memdev_root, find_max_supported_pagesize, &hpsize);
> +        object_child_foreach(memdev_root, find_min_backend_pagesize, &hpsize);
>      }
>      if (hpsize == LONG_MAX) {
>          /* No additional memory regions found ==> Report main RAM page size */
> @@ -1746,8 +1766,24 @@ long qemu_getrampagesize(void)
>  
>      return hpsize;
>  }
> +
> +long qemu_maxrampagesize(void)
> +{
> +    long pagesize = qemu_mempath_getpagesize(mem_path);
> +    Object *memdev_root = object_resolve_path("/objects", NULL);
> +
> +    if (memdev_root) {
> +        object_child_foreach(memdev_root, find_max_backend_pagesize,
> +                             &pagesize);
> +    }
> +    return pagesize;
> +}
>  #else
> -long qemu_getrampagesize(void)
> +long qemu_minrampagesize(void)
> +{
> +    return getpagesize();
> +}
> +long qemu_maxrampagesize(void)
>  {
>      return getpagesize();
>  }
> diff --git a/hw/ppc/spapr_caps.c b/hw/ppc/spapr_caps.c
> index edc5ed0e0c..9b1c10baa6 100644
> --- a/hw/ppc/spapr_caps.c
> +++ b/hw/ppc/spapr_caps.c
> @@ -347,7 +347,7 @@ static void cap_hpt_maxpagesize_apply(SpaprMachineState *spapr,
>          warn_report("Many guests require at least 64kiB hpt-max-page-size");
>      }
>  
> -    spapr_check_pagesize(spapr, qemu_getrampagesize(), errp);
> +    spapr_check_pagesize(spapr, qemu_minrampagesize(), errp);
>  }
>  
>  static bool spapr_pagesize_cb(void *opaque, uint32_t seg_pshift,
> @@ -609,7 +609,7 @@ static SpaprCapabilities default_caps_with_cpu(SpaprMachineState *spapr,
>          uint8_t mps;
>  
>          if (kvmppc_hpt_needs_host_contiguous_pages()) {
> -            mps = ctz64(qemu_getrampagesize());
> +            mps = ctz64(qemu_minrampagesize());
>          } else {
>              mps = 34; /* allow everything up to 16GiB, i.e. everything */
>          }
> diff --git a/hw/s390x/s390-virtio-ccw.c b/hw/s390x/s390-virtio-ccw.c
> index 3be5679657..7e256d3d31 100644
> --- a/hw/s390x/s390-virtio-ccw.c
> +++ b/hw/s390x/s390-virtio-ccw.c
> @@ -188,7 +188,7 @@ static void s390_memory_init(ram_addr_t mem_size)
>       * Configure the maximum page size. As no memory devices were created
>       * yet, this is the page size of initial memory only.
>       */
> -    s390_set_max_pagesize(qemu_getrampagesize(), &local_err);
> +    s390_set_max_pagesize(qemu_maxrampagesize(), &local_err);
>      if (local_err) {
>          error_report_err(local_err);
>          exit(EXIT_FAILURE);
> diff --git a/hw/vfio/spapr.c b/hw/vfio/spapr.c
> index 57fe758e54..96c0ad9d9b 100644
> --- a/hw/vfio/spapr.c
> +++ b/hw/vfio/spapr.c
> @@ -148,7 +148,7 @@ int vfio_spapr_create_window(VFIOContainer *container,
>      uint64_t pagesize = memory_region_iommu_get_min_page_size(iommu_mr);
>      unsigned entries, bits_total, bits_per_level, max_levels;
>      struct vfio_iommu_spapr_tce_create create = { .argsz = sizeof(create) };
> -    long rampagesize = qemu_getrampagesize();
> +    long rampagesize = qemu_minrampagesize();
>  
>      /*
>       * The host might not support the guest supported IOMMU page size,
> diff --git a/include/exec/ram_addr.h b/include/exec/ram_addr.h
> index 9ecd911c3e..139ad79390 100644
> --- a/include/exec/ram_addr.h
> +++ b/include/exec/ram_addr.h
> @@ -73,7 +73,8 @@ static inline unsigned long int ramblock_recv_bitmap_offset(void *host_addr,
>  
>  bool ramblock_is_pmem(RAMBlock *rb);
>  
> -long qemu_getrampagesize(void);
> +long qemu_minrampagesize(void);
> +long qemu_maxrampagesize(void);
>  
>  /**
>   * qemu_ram_alloc_from_file,
> diff --git a/target/ppc/kvm.c b/target/ppc/kvm.c
> index 2427c8ee13..59d92c4275 100644
> --- a/target/ppc/kvm.c
> +++ b/target/ppc/kvm.c
> @@ -2136,7 +2136,7 @@ uint64_t kvmppc_rma_size(uint64_t current_size, unsigned int hash_shift)
>      /* Find the largest hardware supported page size that's less than
>       * or equal to the (logical) backing page size of guest RAM */
>      kvm_get_smmu_info(&info, &error_fatal);
> -    rampagesize = qemu_getrampagesize();
> +    rampagesize = qemu_minrampagesize();
>      best_page_shift = 0;
>  
>      for (i = 0; i < KVM_PPC_PAGE_SIZES_MAX_SZ; i++) {

-- 
David Gibson			| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au	| minimalist, thank you.  NOT _the_ _other_
				| _way_ _around_!
http://www.ozlabs.org/~dgibson

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [Qemu-devel] [PATCH v3 2/2] exec: Introduce qemu_maxrampagesize() and rename qemu_getrampagesize()

[David Gibson]

[-- Attachment #1: Type: text/plain, Size: 6908 bytes --]

On Wed, Apr 17, 2019 at 01:31:43PM +0200, David Hildenbrand wrote:
> Rename qemu_getrampagesize() to qemu_minrampagesize(). While at it,
> properly rename find_max_supported_pagesize() to
> find_min_backend_pagesize().
> 
> s390x is actually interested into the maximum ram pagesize, so
> introduce and use qemu_maxrampagesize().
> 
> Add a TODO, indicating that looking at any mapped memory backends is not
> 100% correct in some cases.
> 
> Signed-off-by: David Hildenbrand <david@redhat.com>

Reviewed-by: David Gibson <david@gibson.dropbear.id.au>

> ---
>  exec.c                     | 44 ++++++++++++++++++++++++++++++++++----
>  hw/ppc/spapr_caps.c        |  4 ++--
>  hw/s390x/s390-virtio-ccw.c |  2 +-
>  hw/vfio/spapr.c            |  2 +-
>  include/exec/ram_addr.h    |  3 ++-
>  target/ppc/kvm.c           |  2 +-
>  6 files changed, 47 insertions(+), 10 deletions(-)
> 
> diff --git a/exec.c b/exec.c
> index 6ab62f4eee..f9b834ef82 100644
> --- a/exec.c
> +++ b/exec.c
> @@ -1687,7 +1687,7 @@ void ram_block_dump(Monitor *mon)
>   * when we actually open and map them.  Iterate over the file
>   * descriptors instead, and use qemu_fd_getpagesize().
>   */
> -static int find_max_supported_pagesize(Object *obj, void *opaque)
> +static int find_min_backend_pagesize(Object *obj, void *opaque)
>  {
>      long *hpsize_min = opaque;
>  
> @@ -1703,7 +1703,27 @@ static int find_max_supported_pagesize(Object *obj, void *opaque)
>      return 0;
>  }
>  
> -long qemu_getrampagesize(void)
> +static int find_max_backend_pagesize(Object *obj, void *opaque)
> +{
> +    long *hpsize_max = opaque;
> +
> +    if (object_dynamic_cast(obj, TYPE_MEMORY_BACKEND)) {
> +        HostMemoryBackend *backend = MEMORY_BACKEND(obj);
> +        long hpsize = host_memory_backend_pagesize(backend);
> +
> +        if (host_memory_backend_is_mapped(backend) && (hpsize > *hpsize_max)) {
> +            *hpsize_max = hpsize;
> +        }
> +    }
> +
> +    return 0;
> +}
> +
> +/*
> + * TODO: We assume right now that all mapped host memory backends are
> + * used as RAM, however some might be used for different purposes.
> + */
> +long qemu_minrampagesize(void)
>  {
>      long hpsize = LONG_MAX;
>      long mainrampagesize;
> @@ -1723,7 +1743,7 @@ long qemu_getrampagesize(void)
>       */
>      memdev_root = object_resolve_path("/objects", NULL);
>      if (memdev_root) {
> -        object_child_foreach(memdev_root, find_max_supported_pagesize, &hpsize);
> +        object_child_foreach(memdev_root, find_min_backend_pagesize, &hpsize);
>      }
>      if (hpsize == LONG_MAX) {
>          /* No additional memory regions found ==> Report main RAM page size */
> @@ -1746,8 +1766,24 @@ long qemu_getrampagesize(void)
>  
>      return hpsize;
>  }
> +
> +long qemu_maxrampagesize(void)
> +{
> +    long pagesize = qemu_mempath_getpagesize(mem_path);
> +    Object *memdev_root = object_resolve_path("/objects", NULL);
> +
> +    if (memdev_root) {
> +        object_child_foreach(memdev_root, find_max_backend_pagesize,
> +                             &pagesize);
> +    }
> +    return pagesize;
> +}
>  #else
> -long qemu_getrampagesize(void)
> +long qemu_minrampagesize(void)
> +{
> +    return getpagesize();
> +}
> +long qemu_maxrampagesize(void)
>  {
>      return getpagesize();
>  }
> diff --git a/hw/ppc/spapr_caps.c b/hw/ppc/spapr_caps.c
> index edc5ed0e0c..9b1c10baa6 100644
> --- a/hw/ppc/spapr_caps.c
> +++ b/hw/ppc/spapr_caps.c
> @@ -347,7 +347,7 @@ static void cap_hpt_maxpagesize_apply(SpaprMachineState *spapr,
>          warn_report("Many guests require at least 64kiB hpt-max-page-size");
>      }
>  
> -    spapr_check_pagesize(spapr, qemu_getrampagesize(), errp);
> +    spapr_check_pagesize(spapr, qemu_minrampagesize(), errp);
>  }
>  
>  static bool spapr_pagesize_cb(void *opaque, uint32_t seg_pshift,
> @@ -609,7 +609,7 @@ static SpaprCapabilities default_caps_with_cpu(SpaprMachineState *spapr,
>          uint8_t mps;
>  
>          if (kvmppc_hpt_needs_host_contiguous_pages()) {
> -            mps = ctz64(qemu_getrampagesize());
> +            mps = ctz64(qemu_minrampagesize());
>          } else {
>              mps = 34; /* allow everything up to 16GiB, i.e. everything */
>          }
> diff --git a/hw/s390x/s390-virtio-ccw.c b/hw/s390x/s390-virtio-ccw.c
> index 3be5679657..7e256d3d31 100644
> --- a/hw/s390x/s390-virtio-ccw.c
> +++ b/hw/s390x/s390-virtio-ccw.c
> @@ -188,7 +188,7 @@ static void s390_memory_init(ram_addr_t mem_size)
>       * Configure the maximum page size. As no memory devices were created
>       * yet, this is the page size of initial memory only.
>       */
> -    s390_set_max_pagesize(qemu_getrampagesize(), &local_err);
> +    s390_set_max_pagesize(qemu_maxrampagesize(), &local_err);
>      if (local_err) {
>          error_report_err(local_err);
>          exit(EXIT_FAILURE);
> diff --git a/hw/vfio/spapr.c b/hw/vfio/spapr.c
> index 57fe758e54..96c0ad9d9b 100644
> --- a/hw/vfio/spapr.c
> +++ b/hw/vfio/spapr.c
> @@ -148,7 +148,7 @@ int vfio_spapr_create_window(VFIOContainer *container,
>      uint64_t pagesize = memory_region_iommu_get_min_page_size(iommu_mr);
>      unsigned entries, bits_total, bits_per_level, max_levels;
>      struct vfio_iommu_spapr_tce_create create = { .argsz = sizeof(create) };
> -    long rampagesize = qemu_getrampagesize();
> +    long rampagesize = qemu_minrampagesize();
>  
>      /*
>       * The host might not support the guest supported IOMMU page size,
> diff --git a/include/exec/ram_addr.h b/include/exec/ram_addr.h
> index 9ecd911c3e..139ad79390 100644
> --- a/include/exec/ram_addr.h
> +++ b/include/exec/ram_addr.h
> @@ -73,7 +73,8 @@ static inline unsigned long int ramblock_recv_bitmap_offset(void *host_addr,
>  
>  bool ramblock_is_pmem(RAMBlock *rb);
>  
> -long qemu_getrampagesize(void);
> +long qemu_minrampagesize(void);
> +long qemu_maxrampagesize(void);
>  
>  /**
>   * qemu_ram_alloc_from_file,
> diff --git a/target/ppc/kvm.c b/target/ppc/kvm.c
> index 2427c8ee13..59d92c4275 100644
> --- a/target/ppc/kvm.c
> +++ b/target/ppc/kvm.c
> @@ -2136,7 +2136,7 @@ uint64_t kvmppc_rma_size(uint64_t current_size, unsigned int hash_shift)
>      /* Find the largest hardware supported page size that's less than
>       * or equal to the (logical) backing page size of guest RAM */
>      kvm_get_smmu_info(&info, &error_fatal);
> -    rampagesize = qemu_getrampagesize();
> +    rampagesize = qemu_minrampagesize();
>      best_page_shift = 0;
>  
>      for (i = 0; i < KVM_PPC_PAGE_SIZES_MAX_SZ; i++) {

-- 
David Gibson			| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au	| minimalist, thank you.  NOT _the_ _other_
				| _way_ _around_!
http://www.ozlabs.org/~dgibson

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [Qemu-devel] [PATCH v3 2/2] exec: Introduce qemu_maxrampagesize() and rename qemu_getrampagesize()

[Igor Mammedov]

On Wed, 17 Apr 2019 13:31:43 +0200
David Hildenbrand <david@redhat.com> wrote:

> Rename qemu_getrampagesize() to qemu_minrampagesize(). While at it,
> properly rename find_max_supported_pagesize() to
> find_min_backend_pagesize().
> 
> s390x is actually interested into the maximum ram pagesize, so
> introduce and use qemu_maxrampagesize().
> 
> Add a TODO, indicating that looking at any mapped memory backends is not
> 100% correct in some cases.
> 
> Signed-off-by: David Hildenbrand <david@redhat.com>

Reviewed-by: Igor Mammedov <imammedo@redhat.com>

> ---
>  exec.c                     | 44 ++++++++++++++++++++++++++++++++++----
>  hw/ppc/spapr_caps.c        |  4 ++--
>  hw/s390x/s390-virtio-ccw.c |  2 +-
>  hw/vfio/spapr.c            |  2 +-
>  include/exec/ram_addr.h    |  3 ++-
>  target/ppc/kvm.c           |  2 +-
>  6 files changed, 47 insertions(+), 10 deletions(-)
> 
> diff --git a/exec.c b/exec.c
> index 6ab62f4eee..f9b834ef82 100644
> --- a/exec.c
> +++ b/exec.c
> @@ -1687,7 +1687,7 @@ void ram_block_dump(Monitor *mon)
>   * when we actually open and map them.  Iterate over the file
>   * descriptors instead, and use qemu_fd_getpagesize().
>   */
> -static int find_max_supported_pagesize(Object *obj, void *opaque)
> +static int find_min_backend_pagesize(Object *obj, void *opaque)
>  {
>      long *hpsize_min = opaque;
>  
> @@ -1703,7 +1703,27 @@ static int find_max_supported_pagesize(Object *obj, void *opaque)
>      return 0;
>  }
>  
> -long qemu_getrampagesize(void)
> +static int find_max_backend_pagesize(Object *obj, void *opaque)
> +{
> +    long *hpsize_max = opaque;
> +
> +    if (object_dynamic_cast(obj, TYPE_MEMORY_BACKEND)) {
> +        HostMemoryBackend *backend = MEMORY_BACKEND(obj);
> +        long hpsize = host_memory_backend_pagesize(backend);
> +
> +        if (host_memory_backend_is_mapped(backend) && (hpsize > *hpsize_max)) {
> +            *hpsize_max = hpsize;
> +        }
> +    }
> +
> +    return 0;
> +}
> +
> +/*
> + * TODO: We assume right now that all mapped host memory backends are
> + * used as RAM, however some might be used for different purposes.
> + */
> +long qemu_minrampagesize(void)
>  {
>      long hpsize = LONG_MAX;
>      long mainrampagesize;
> @@ -1723,7 +1743,7 @@ long qemu_getrampagesize(void)
>       */
>      memdev_root = object_resolve_path("/objects", NULL);
>      if (memdev_root) {
> -        object_child_foreach(memdev_root, find_max_supported_pagesize, &hpsize);
> +        object_child_foreach(memdev_root, find_min_backend_pagesize, &hpsize);
>      }
>      if (hpsize == LONG_MAX) {
>          /* No additional memory regions found ==> Report main RAM page size */
> @@ -1746,8 +1766,24 @@ long qemu_getrampagesize(void)
>  
>      return hpsize;
>  }
> +
> +long qemu_maxrampagesize(void)
> +{
> +    long pagesize = qemu_mempath_getpagesize(mem_path);
> +    Object *memdev_root = object_resolve_path("/objects", NULL);
> +
> +    if (memdev_root) {
> +        object_child_foreach(memdev_root, find_max_backend_pagesize,
> +                             &pagesize);
> +    }
> +    return pagesize;
> +}
>  #else
> -long qemu_getrampagesize(void)
> +long qemu_minrampagesize(void)
> +{
> +    return getpagesize();
> +}
> +long qemu_maxrampagesize(void)
>  {
>      return getpagesize();
>  }
> diff --git a/hw/ppc/spapr_caps.c b/hw/ppc/spapr_caps.c
> index edc5ed0e0c..9b1c10baa6 100644
> --- a/hw/ppc/spapr_caps.c
> +++ b/hw/ppc/spapr_caps.c
> @@ -347,7 +347,7 @@ static void cap_hpt_maxpagesize_apply(SpaprMachineState *spapr,
>          warn_report("Many guests require at least 64kiB hpt-max-page-size");
>      }
>  
> -    spapr_check_pagesize(spapr, qemu_getrampagesize(), errp);
> +    spapr_check_pagesize(spapr, qemu_minrampagesize(), errp);
>  }
>  
>  static bool spapr_pagesize_cb(void *opaque, uint32_t seg_pshift,
> @@ -609,7 +609,7 @@ static SpaprCapabilities default_caps_with_cpu(SpaprMachineState *spapr,
>          uint8_t mps;
>  
>          if (kvmppc_hpt_needs_host_contiguous_pages()) {
> -            mps = ctz64(qemu_getrampagesize());
> +            mps = ctz64(qemu_minrampagesize());
>          } else {
>              mps = 34; /* allow everything up to 16GiB, i.e. everything */
>          }
> diff --git a/hw/s390x/s390-virtio-ccw.c b/hw/s390x/s390-virtio-ccw.c
> index 3be5679657..7e256d3d31 100644
> --- a/hw/s390x/s390-virtio-ccw.c
> +++ b/hw/s390x/s390-virtio-ccw.c
> @@ -188,7 +188,7 @@ static void s390_memory_init(ram_addr_t mem_size)
>       * Configure the maximum page size. As no memory devices were created
>       * yet, this is the page size of initial memory only.
>       */
> -    s390_set_max_pagesize(qemu_getrampagesize(), &local_err);
> +    s390_set_max_pagesize(qemu_maxrampagesize(), &local_err);
>      if (local_err) {
>          error_report_err(local_err);
>          exit(EXIT_FAILURE);
> diff --git a/hw/vfio/spapr.c b/hw/vfio/spapr.c
> index 57fe758e54..96c0ad9d9b 100644
> --- a/hw/vfio/spapr.c
> +++ b/hw/vfio/spapr.c
> @@ -148,7 +148,7 @@ int vfio_spapr_create_window(VFIOContainer *container,
>      uint64_t pagesize = memory_region_iommu_get_min_page_size(iommu_mr);
>      unsigned entries, bits_total, bits_per_level, max_levels;
>      struct vfio_iommu_spapr_tce_create create = { .argsz = sizeof(create) };
> -    long rampagesize = qemu_getrampagesize();
> +    long rampagesize = qemu_minrampagesize();
>  
>      /*
>       * The host might not support the guest supported IOMMU page size,
> diff --git a/include/exec/ram_addr.h b/include/exec/ram_addr.h
> index 9ecd911c3e..139ad79390 100644
> --- a/include/exec/ram_addr.h
> +++ b/include/exec/ram_addr.h
> @@ -73,7 +73,8 @@ static inline unsigned long int ramblock_recv_bitmap_offset(void *host_addr,
>  
>  bool ramblock_is_pmem(RAMBlock *rb);
>  
> -long qemu_getrampagesize(void);
> +long qemu_minrampagesize(void);
> +long qemu_maxrampagesize(void);
>  
>  /**
>   * qemu_ram_alloc_from_file,
> diff --git a/target/ppc/kvm.c b/target/ppc/kvm.c
> index 2427c8ee13..59d92c4275 100644
> --- a/target/ppc/kvm.c
> +++ b/target/ppc/kvm.c
> @@ -2136,7 +2136,7 @@ uint64_t kvmppc_rma_size(uint64_t current_size, unsigned int hash_shift)
>      /* Find the largest hardware supported page size that's less than
>       * or equal to the (logical) backing page size of guest RAM */
>      kvm_get_smmu_info(&info, &error_fatal);
> -    rampagesize = qemu_getrampagesize();
> +    rampagesize = qemu_minrampagesize();
>      best_page_shift = 0;
>  
>      for (i = 0; i < KVM_PPC_PAGE_SIZES_MAX_SZ; i++) {

Re: [Qemu-devel] [PATCH v3 2/2] exec: Introduce qemu_maxrampagesize() and rename qemu_getrampagesize()

[Igor Mammedov]

On Wed, 17 Apr 2019 13:31:43 +0200
David Hildenbrand <david@redhat.com> wrote:

> Rename qemu_getrampagesize() to qemu_minrampagesize(). While at it,
> properly rename find_max_supported_pagesize() to
> find_min_backend_pagesize().
> 
> s390x is actually interested into the maximum ram pagesize, so
> introduce and use qemu_maxrampagesize().
> 
> Add a TODO, indicating that looking at any mapped memory backends is not
> 100% correct in some cases.
> 
> Signed-off-by: David Hildenbrand <david@redhat.com>

Reviewed-by: Igor Mammedov <imammedo@redhat.com>

> ---
>  exec.c                     | 44 ++++++++++++++++++++++++++++++++++----
>  hw/ppc/spapr_caps.c        |  4 ++--
>  hw/s390x/s390-virtio-ccw.c |  2 +-
>  hw/vfio/spapr.c            |  2 +-
>  include/exec/ram_addr.h    |  3 ++-
>  target/ppc/kvm.c           |  2 +-
>  6 files changed, 47 insertions(+), 10 deletions(-)
> 
> diff --git a/exec.c b/exec.c
> index 6ab62f4eee..f9b834ef82 100644
> --- a/exec.c
> +++ b/exec.c
> @@ -1687,7 +1687,7 @@ void ram_block_dump(Monitor *mon)
>   * when we actually open and map them.  Iterate over the file
>   * descriptors instead, and use qemu_fd_getpagesize().
>   */
> -static int find_max_supported_pagesize(Object *obj, void *opaque)
> +static int find_min_backend_pagesize(Object *obj, void *opaque)
>  {
>      long *hpsize_min = opaque;
>  
> @@ -1703,7 +1703,27 @@ static int find_max_supported_pagesize(Object *obj, void *opaque)
>      return 0;
>  }
>  
> -long qemu_getrampagesize(void)
> +static int find_max_backend_pagesize(Object *obj, void *opaque)
> +{
> +    long *hpsize_max = opaque;
> +
> +    if (object_dynamic_cast(obj, TYPE_MEMORY_BACKEND)) {
> +        HostMemoryBackend *backend = MEMORY_BACKEND(obj);
> +        long hpsize = host_memory_backend_pagesize(backend);
> +
> +        if (host_memory_backend_is_mapped(backend) && (hpsize > *hpsize_max)) {
> +            *hpsize_max = hpsize;
> +        }
> +    }
> +
> +    return 0;
> +}
> +
> +/*
> + * TODO: We assume right now that all mapped host memory backends are
> + * used as RAM, however some might be used for different purposes.
> + */
> +long qemu_minrampagesize(void)
>  {
>      long hpsize = LONG_MAX;
>      long mainrampagesize;
> @@ -1723,7 +1743,7 @@ long qemu_getrampagesize(void)
>       */
>      memdev_root = object_resolve_path("/objects", NULL);
>      if (memdev_root) {
> -        object_child_foreach(memdev_root, find_max_supported_pagesize, &hpsize);
> +        object_child_foreach(memdev_root, find_min_backend_pagesize, &hpsize);
>      }
>      if (hpsize == LONG_MAX) {
>          /* No additional memory regions found ==> Report main RAM page size */
> @@ -1746,8 +1766,24 @@ long qemu_getrampagesize(void)
>  
>      return hpsize;
>  }
> +
> +long qemu_maxrampagesize(void)
> +{
> +    long pagesize = qemu_mempath_getpagesize(mem_path);
> +    Object *memdev_root = object_resolve_path("/objects", NULL);
> +
> +    if (memdev_root) {
> +        object_child_foreach(memdev_root, find_max_backend_pagesize,
> +                             &pagesize);
> +    }
> +    return pagesize;
> +}
>  #else
> -long qemu_getrampagesize(void)
> +long qemu_minrampagesize(void)
> +{
> +    return getpagesize();
> +}
> +long qemu_maxrampagesize(void)
>  {
>      return getpagesize();
>  }
> diff --git a/hw/ppc/spapr_caps.c b/hw/ppc/spapr_caps.c
> index edc5ed0e0c..9b1c10baa6 100644
> --- a/hw/ppc/spapr_caps.c
> +++ b/hw/ppc/spapr_caps.c
> @@ -347,7 +347,7 @@ static void cap_hpt_maxpagesize_apply(SpaprMachineState *spapr,
>          warn_report("Many guests require at least 64kiB hpt-max-page-size");
>      }
>  
> -    spapr_check_pagesize(spapr, qemu_getrampagesize(), errp);
> +    spapr_check_pagesize(spapr, qemu_minrampagesize(), errp);
>  }
>  
>  static bool spapr_pagesize_cb(void *opaque, uint32_t seg_pshift,
> @@ -609,7 +609,7 @@ static SpaprCapabilities default_caps_with_cpu(SpaprMachineState *spapr,
>          uint8_t mps;
>  
>          if (kvmppc_hpt_needs_host_contiguous_pages()) {
> -            mps = ctz64(qemu_getrampagesize());
> +            mps = ctz64(qemu_minrampagesize());
>          } else {
>              mps = 34; /* allow everything up to 16GiB, i.e. everything */
>          }
> diff --git a/hw/s390x/s390-virtio-ccw.c b/hw/s390x/s390-virtio-ccw.c
> index 3be5679657..7e256d3d31 100644
> --- a/hw/s390x/s390-virtio-ccw.c
> +++ b/hw/s390x/s390-virtio-ccw.c
> @@ -188,7 +188,7 @@ static void s390_memory_init(ram_addr_t mem_size)
>       * Configure the maximum page size. As no memory devices were created
>       * yet, this is the page size of initial memory only.
>       */
> -    s390_set_max_pagesize(qemu_getrampagesize(), &local_err);
> +    s390_set_max_pagesize(qemu_maxrampagesize(), &local_err);
>      if (local_err) {
>          error_report_err(local_err);
>          exit(EXIT_FAILURE);
> diff --git a/hw/vfio/spapr.c b/hw/vfio/spapr.c
> index 57fe758e54..96c0ad9d9b 100644
> --- a/hw/vfio/spapr.c
> +++ b/hw/vfio/spapr.c
> @@ -148,7 +148,7 @@ int vfio_spapr_create_window(VFIOContainer *container,
>      uint64_t pagesize = memory_region_iommu_get_min_page_size(iommu_mr);
>      unsigned entries, bits_total, bits_per_level, max_levels;
>      struct vfio_iommu_spapr_tce_create create = { .argsz = sizeof(create) };
> -    long rampagesize = qemu_getrampagesize();
> +    long rampagesize = qemu_minrampagesize();
>  
>      /*
>       * The host might not support the guest supported IOMMU page size,
> diff --git a/include/exec/ram_addr.h b/include/exec/ram_addr.h
> index 9ecd911c3e..139ad79390 100644
> --- a/include/exec/ram_addr.h
> +++ b/include/exec/ram_addr.h
> @@ -73,7 +73,8 @@ static inline unsigned long int ramblock_recv_bitmap_offset(void *host_addr,
>  
>  bool ramblock_is_pmem(RAMBlock *rb);
>  
> -long qemu_getrampagesize(void);
> +long qemu_minrampagesize(void);
> +long qemu_maxrampagesize(void);
>  
>  /**
>   * qemu_ram_alloc_from_file,
> diff --git a/target/ppc/kvm.c b/target/ppc/kvm.c
> index 2427c8ee13..59d92c4275 100644
> --- a/target/ppc/kvm.c
> +++ b/target/ppc/kvm.c
> @@ -2136,7 +2136,7 @@ uint64_t kvmppc_rma_size(uint64_t current_size, unsigned int hash_shift)
>      /* Find the largest hardware supported page size that's less than
>       * or equal to the (logical) backing page size of guest RAM */
>      kvm_get_smmu_info(&info, &error_fatal);
> -    rampagesize = qemu_getrampagesize();
> +    rampagesize = qemu_minrampagesize();
>      best_page_shift = 0;
>  
>      for (i = 0; i < KVM_PPC_PAGE_SIZES_MAX_SZ; i++) {

Re: [Qemu-devel] [PATCH v3 0/2] s390x/kvm: Properly detect page size of initial memory

[Cornelia Huck]

On Wed, 17 Apr 2019 13:31:41 +0200
David Hildenbrand <david@redhat.com> wrote:

> While the detection on s390x of the maximum ram pagesize works by pure luck,
> we should try to do better. Especially, once "-mem-path" is replaced
> by memory backends, the current approach will break.
> 
> Configure the maximum page size after configuring initital memory. Provide
> a function to detect the maximum ram pagesize
> 
> I think this can be deferred to 4.1 as I don't think anything is actually
> broken right now.

I think it is certainly 4.1 at that point in time ;)

Who is going to queue this? If not me, I'll try to find some time to
review this next week.

> 
> v2 -> v3:
> - Fix compile error
> - Tested that hugepages correctly result in an error with
>   "modprobe kvm hpage=0", but not with "modprobe kvm hpage=1".
> 
> v1 -> v2:
> - Add a better comment describing that only initial memory is considered
> - Add "exec: Introduce qemu_getmaxrampagesize() and rename
>   qemu_getrampagesize()"
> 
> David Hildenbrand (2):
>   s390x/kvm: Configure page size after memory has actually been
>     initialized
>   exec: Introduce qemu_maxrampagesize() and rename qemu_getrampagesize()
> 
>  exec.c                     | 44 ++++++++++++++++++++++++++++++++++----
>  hw/ppc/spapr_caps.c        |  4 ++--
>  hw/s390x/s390-virtio-ccw.c | 12 +++++++++++
>  hw/vfio/spapr.c            |  2 +-
>  include/exec/ram_addr.h    |  3 ++-
>  target/ppc/kvm.c           |  2 +-
>  target/s390x/cpu.c         |  7 ++++++
>  target/s390x/cpu.h         |  1 +
>  target/s390x/kvm-stub.c    |  4 ++++
>  target/s390x/kvm.c         | 35 ++++++++++++------------------
>  target/s390x/kvm_s390x.h   |  1 +
>  11 files changed, 85 insertions(+), 30 deletions(-)
> 

Re: [Qemu-devel] [PATCH v3 0/2] s390x/kvm: Properly detect page size of initial memory

[Cornelia Huck]

On Wed, 17 Apr 2019 13:31:41 +0200
David Hildenbrand <david@redhat.com> wrote:

> While the detection on s390x of the maximum ram pagesize works by pure luck,
> we should try to do better. Especially, once "-mem-path" is replaced
> by memory backends, the current approach will break.
> 
> Configure the maximum page size after configuring initital memory. Provide
> a function to detect the maximum ram pagesize
> 
> I think this can be deferred to 4.1 as I don't think anything is actually
> broken right now.

I think it is certainly 4.1 at that point in time ;)

Who is going to queue this? If not me, I'll try to find some time to
review this next week.

> 
> v2 -> v3:
> - Fix compile error
> - Tested that hugepages correctly result in an error with
>   "modprobe kvm hpage=0", but not with "modprobe kvm hpage=1".
> 
> v1 -> v2:
> - Add a better comment describing that only initial memory is considered
> - Add "exec: Introduce qemu_getmaxrampagesize() and rename
>   qemu_getrampagesize()"
> 
> David Hildenbrand (2):
>   s390x/kvm: Configure page size after memory has actually been
>     initialized
>   exec: Introduce qemu_maxrampagesize() and rename qemu_getrampagesize()
> 
>  exec.c                     | 44 ++++++++++++++++++++++++++++++++++----
>  hw/ppc/spapr_caps.c        |  4 ++--
>  hw/s390x/s390-virtio-ccw.c | 12 +++++++++++
>  hw/vfio/spapr.c            |  2 +-
>  include/exec/ram_addr.h    |  3 ++-
>  target/ppc/kvm.c           |  2 +-
>  target/s390x/cpu.c         |  7 ++++++
>  target/s390x/cpu.h         |  1 +
>  target/s390x/kvm-stub.c    |  4 ++++
>  target/s390x/kvm.c         | 35 ++++++++++++------------------
>  target/s390x/kvm_s390x.h   |  1 +
>  11 files changed, 85 insertions(+), 30 deletions(-)
> 

Re: [Qemu-devel] [PATCH v3 2/2] exec: Introduce qemu_maxrampagesize() and rename qemu_getrampagesize()

[Cornelia Huck]

On Wed, 17 Apr 2019 13:31:43 +0200
David Hildenbrand <david@redhat.com> wrote:

> Rename qemu_getrampagesize() to qemu_minrampagesize(). While at it,
> properly rename find_max_supported_pagesize() to
> find_min_backend_pagesize().
> 
> s390x is actually interested into the maximum ram pagesize, so
> introduce and use qemu_maxrampagesize().
> 
> Add a TODO, indicating that looking at any mapped memory backends is not
> 100% correct in some cases.
> 
> Signed-off-by: David Hildenbrand <david@redhat.com>
> ---
>  exec.c                     | 44 ++++++++++++++++++++++++++++++++++----
>  hw/ppc/spapr_caps.c        |  4 ++--
>  hw/s390x/s390-virtio-ccw.c |  2 +-
>  hw/vfio/spapr.c            |  2 +-
>  include/exec/ram_addr.h    |  3 ++-
>  target/ppc/kvm.c           |  2 +-
>  6 files changed, 47 insertions(+), 10 deletions(-)

This now has two R-bs, so unless somebody speaks up against it, I'll
just take this through the s390x tree together with patch 1.

Re: [Qemu-devel] [PATCH v3 2/2] exec: Introduce qemu_maxrampagesize() and rename qemu_getrampagesize()

[Cornelia Huck]

On Wed, 17 Apr 2019 13:31:43 +0200
David Hildenbrand <david@redhat.com> wrote:

> Rename qemu_getrampagesize() to qemu_minrampagesize(). While at it,
> properly rename find_max_supported_pagesize() to
> find_min_backend_pagesize().
> 
> s390x is actually interested into the maximum ram pagesize, so
> introduce and use qemu_maxrampagesize().
> 
> Add a TODO, indicating that looking at any mapped memory backends is not
> 100% correct in some cases.
> 
> Signed-off-by: David Hildenbrand <david@redhat.com>
> ---
>  exec.c                     | 44 ++++++++++++++++++++++++++++++++++----
>  hw/ppc/spapr_caps.c        |  4 ++--
>  hw/s390x/s390-virtio-ccw.c |  2 +-
>  hw/vfio/spapr.c            |  2 +-
>  include/exec/ram_addr.h    |  3 ++-
>  target/ppc/kvm.c           |  2 +-
>  6 files changed, 47 insertions(+), 10 deletions(-)

This now has two R-bs, so unless somebody speaks up against it, I'll
just take this through the s390x tree together with patch 1.

Re: [Qemu-devel] [PATCH v3 0/2] s390x/kvm: Properly detect page size of initial memory

[Cornelia Huck]

On Wed, 17 Apr 2019 13:31:41 +0200
David Hildenbrand <david@redhat.com> wrote:

> While the detection on s390x of the maximum ram pagesize works by pure luck,
> we should try to do better. Especially, once "-mem-path" is replaced
> by memory backends, the current approach will break.
> 
> Configure the maximum page size after configuring initital memory. Provide
> a function to detect the maximum ram pagesize
> 
> I think this can be deferred to 4.1 as I don't think anything is actually
> broken right now.
> 
> v2 -> v3:
> - Fix compile error
> - Tested that hugepages correctly result in an error with
>   "modprobe kvm hpage=0", but not with "modprobe kvm hpage=1".
> 
> v1 -> v2:
> - Add a better comment describing that only initial memory is considered
> - Add "exec: Introduce qemu_getmaxrampagesize() and rename
>   qemu_getrampagesize()"
> 
> David Hildenbrand (2):
>   s390x/kvm: Configure page size after memory has actually been
>     initialized
>   exec: Introduce qemu_maxrampagesize() and rename qemu_getrampagesize()
> 
>  exec.c                     | 44 ++++++++++++++++++++++++++++++++++----
>  hw/ppc/spapr_caps.c        |  4 ++--
>  hw/s390x/s390-virtio-ccw.c | 12 +++++++++++
>  hw/vfio/spapr.c            |  2 +-
>  include/exec/ram_addr.h    |  3 ++-
>  target/ppc/kvm.c           |  2 +-
>  target/s390x/cpu.c         |  7 ++++++
>  target/s390x/cpu.h         |  1 +
>  target/s390x/kvm-stub.c    |  4 ++++
>  target/s390x/kvm.c         | 35 ++++++++++++------------------
>  target/s390x/kvm_s390x.h   |  1 +
>  11 files changed, 85 insertions(+), 30 deletions(-)
> 

Thanks, applied.

Re: [Qemu-devel] [PATCH v3 0/2] s390x/kvm: Properly detect page size of initial memory

[Cornelia Huck]

On Wed, 17 Apr 2019 13:31:41 +0200
David Hildenbrand <david@redhat.com> wrote:

> While the detection on s390x of the maximum ram pagesize works by pure luck,
> we should try to do better. Especially, once "-mem-path" is replaced
> by memory backends, the current approach will break.
> 
> Configure the maximum page size after configuring initital memory. Provide
> a function to detect the maximum ram pagesize
> 
> I think this can be deferred to 4.1 as I don't think anything is actually
> broken right now.
> 
> v2 -> v3:
> - Fix compile error
> - Tested that hugepages correctly result in an error with
>   "modprobe kvm hpage=0", but not with "modprobe kvm hpage=1".
> 
> v1 -> v2:
> - Add a better comment describing that only initial memory is considered
> - Add "exec: Introduce qemu_getmaxrampagesize() and rename
>   qemu_getrampagesize()"
> 
> David Hildenbrand (2):
>   s390x/kvm: Configure page size after memory has actually been
>     initialized
>   exec: Introduce qemu_maxrampagesize() and rename qemu_getrampagesize()
> 
>  exec.c                     | 44 ++++++++++++++++++++++++++++++++++----
>  hw/ppc/spapr_caps.c        |  4 ++--
>  hw/s390x/s390-virtio-ccw.c | 12 +++++++++++
>  hw/vfio/spapr.c            |  2 +-
>  include/exec/ram_addr.h    |  3 ++-
>  target/ppc/kvm.c           |  2 +-
>  target/s390x/cpu.c         |  7 ++++++
>  target/s390x/cpu.h         |  1 +
>  target/s390x/kvm-stub.c    |  4 ++++
>  target/s390x/kvm.c         | 35 ++++++++++++------------------
>  target/s390x/kvm_s390x.h   |  1 +
>  11 files changed, 85 insertions(+), 30 deletions(-)
> 

Thanks, applied.