[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [PATCH] hostmem: Don't report pmem attribute if unsupported

From: Daniel Henrique Barboza
Subject: Re: [PATCH] hostmem: Don't report pmem attribute if unsupported
Date: Tue, 19 Jan 2021 13:55:02 -0300
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101 Thunderbird/78.6.0

On 1/19/21 9:35 AM, Michal Privoznik wrote:
When management applications (like Libvirt) want to check whether
memory-backend-file.pmem is supported they can list object
properties using 'qom-list-properties'. However, 'pmem' is
declared always (and thus reported always) and only at runtime
QEMU errors out if it was built without libpmem (and thus can not
guarantee write persistence). This is suboptimal since we have
ability to declare attributes at compile time.

Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1915216
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>

Testing it on a host that doesn't have libpmem support:

- without this patch:

{"execute": "qom-list-properties", "arguments": {"typename": 
{"return": [
   {"name": "pmem", "type": "bool"},

- with this patch there is no 'pmem' support reported:

{"execute": "qom-list-properties", "arguments": {"typename": 
{"return": [
    {"name": "type", "type": "string"},
    {"name": "policy", "description": "Set the NUMA policy", "type": 
    {"name": "prealloc-threads", "description": "Number of CPU threads to use for prealloc", 
"type": "int"},
    {"name": "dump", "description": "Set to 'off' to exclude from core dump", "type": 
    {"name": "share", "description": "Mark the memory as private to QEMU or shared", 
"type": "bool"},
    {"name": "prealloc", "description": "Preallocate memory", "type": "bool"},
    {"name": "size", "description": "Size of the memory region (ex: 500M)", "type": 
    {"name": "x-use-canonical-path-for-ramblock-id", "type": "bool"},
    {"name": "host-nodes", "description": "Binds memory to the list of NUMA host nodes", 
"type": "int"},
    {"name": "merge", "description": "Mark memory as mergeable", "type": 
    {"name": "align", "type": "int"},
    {"name": "mem-path", "type": "string"},
    {"name": "discard-data", "type": "bool"}

Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Tested-by: Daniel Henrique Barboza <danielhb413@gmail.com>

  backends/hostmem-file.c | 13 ++++---------
  1 file changed, 4 insertions(+), 9 deletions(-)

diff --git a/backends/hostmem-file.c b/backends/hostmem-file.c
index 40e1e5b3e3..7e30eb5985 100644
--- a/backends/hostmem-file.c
+++ b/backends/hostmem-file.c
@@ -123,6 +123,7 @@ static void file_memory_backend_set_align(Object *o, 
Visitor *v,
      fb->align = val;
  static bool file_memory_backend_get_pmem(Object *o, Error **errp)
      return MEMORY_BACKEND_FILE(o)->is_pmem;
@@ -139,17 +140,9 @@ static void file_memory_backend_set_pmem(Object *o, bool 
value, Error **errp)
-    if (value) {
-        error_setg(errp, "Lack of libpmem support while setting the 'pmem=on'"
-                   " of %s. We can't ensure data persistence.",
-                   object_get_typename(o));
-        return;
-    }
      fb->is_pmem = value;
+#endif /* CONFIG_LIBPMEM */
static void file_backend_unparent(Object *obj)
@@ -180,8 +173,10 @@ file_backend_class_init(ObjectClass *oc, void *data)
          NULL, NULL);
      object_class_property_add_bool(oc, "pmem",
          file_memory_backend_get_pmem, file_memory_backend_set_pmem);
static void file_backend_instance_finalize(Object *o)

reply via email to

[Prev in Thread] Current Thread [Next in Thread]