On Mon, 29 Jun 2015 19:11:30 +0530
Bharata B Rao <bhar...@linux.vnet.ibm.com> wrote:
> On Mon, Jun 29, 2015 at 02:08:20PM +0200, Igor Mammedov wrote:
> > On Mon, 29 Jun 2015 13:50:25 +0530
> > Bharata B Rao <bhar...@linux.vnet.ibm.com> wrote:
> >
> > > Start storing the (start_addr, end_addr) of the pc-dimm memory
> > > in corresponding numa_info[node] so that this information can be used
> > > to lookup node by address.
> > >
> > > Signed-off-by: Bharata B Rao <bhar...@linux.vnet.ibm.com>
> > Reviewed-by: Igor Mammedov <imamm...@redhat.com>
> >
> > > ---
> > > hw/mem/pc-dimm.c | 4 ++++
> > > include/sysemu/numa.h | 10 ++++++++++
> > > numa.c | 26 ++++++++++++++++++++++++++
> > > 3 files changed, 40 insertions(+)
> > >
> > > diff --git a/hw/mem/pc-dimm.c b/hw/mem/pc-dimm.c
> > > index 98971b7..bb04862 100644
> > > --- a/hw/mem/pc-dimm.c
> > > +++ b/hw/mem/pc-dimm.c
> > > @@ -97,6 +97,7 @@ void pc_dimm_memory_plug(DeviceState *dev,
> > > MemoryHotplugState *hpms,
> > >
> > > memory_region_add_subregion(&hpms->mr, addr - hpms->base, mr);
> > > vmstate_register_ram(mr, dev);
> > > + numa_set_mem_node_id(addr, memory_region_size(mr), dimm->node);
> > >
> > > out:
> > > error_propagate(errp, local_err);
> > > @@ -105,6 +106,9 @@ out:
> > > void pc_dimm_memory_unplug(DeviceState *dev, MemoryHotplugState *hpms,
> > > MemoryRegion *mr)
> > > {
> > > + PCDIMMDevice *dimm = PC_DIMM(dev);
> > > +
> > > + numa_unset_mem_node_id(dimm->addr, memory_region_size(mr),
> > > dimm->node);
> > > memory_region_del_subregion(&hpms->mr, mr);
> > > vmstate_unregister_ram(mr, dev);
> > > }
> > > diff --git a/include/sysemu/numa.h b/include/sysemu/numa.h
> > > index 6523b4d..7176364 100644
> > > --- a/include/sysemu/numa.h
> > > +++ b/include/sysemu/numa.h
> > > @@ -10,16 +10,26 @@
> > >
> > > extern int nb_numa_nodes; /* Number of NUMA nodes */
> > >
> > > +struct numa_addr_range {
> > > + ram_addr_t mem_start;
> > > + ram_addr_t mem_end;
> > > + QLIST_ENTRY(numa_addr_range) entry;
> > > +};
> > > +
> > > typedef struct node_info {
> > > uint64_t node_mem;
> > > DECLARE_BITMAP(node_cpu, MAX_CPUMASK_BITS);
> > > struct HostMemoryBackend *node_memdev;
> > > bool present;
> > > + QLIST_HEAD(, numa_addr_range) addr; /* List to store address ranges
> > > */
> > > } NodeInfo;
> > > +
> > > extern NodeInfo numa_info[MAX_NODES];
> > > void parse_numa_opts(MachineClass *mc);
> > > void numa_post_machine_init(void);
> > > void query_numa_node_mem(uint64_t node_mem[]);
> > > extern QemuOptsList qemu_numa_opts;
> > > +void numa_set_mem_node_id(ram_addr_t addr, uint64_t size, uint32_t node);
> > > +void numa_unset_mem_node_id(ram_addr_t addr, uint64_t size, uint32_t
> > > node);
> > >
> > > #endif
> > > diff --git a/numa.c b/numa.c
> > > index 91fc6c1..116d1fb 100644
> > > --- a/numa.c
> > > +++ b/numa.c
> > > @@ -52,6 +52,28 @@ static int max_numa_nodeid; /* Highest specified NUMA
> > > node ID, plus one.
> > > int nb_numa_nodes;
> > > NodeInfo numa_info[MAX_NODES];
> > >
> > > +void numa_set_mem_node_id(ram_addr_t addr, uint64_t size, uint32_t node)
> > > +{
> > > + struct numa_addr_range *range = g_malloc0(sizeof(*range));
> > > +
> > > + range->mem_start = addr;
> > > + range->mem_end = addr + size - 1;
> > nit:
> > as a patch on top of it, add asserts that check for overflow, pls
>
> You suggested g_assert(size) in the previous version.
>
> However size can be zero when this API is called for boot time memory
> and I have taken care of that in the next patch (5/6).
>
> And for pc-dimm memory, the size can never be zero.
>
> So do you still think overflow is possible ?
make build on 32-bit host and theoretically with ram_addr_t==32bits + size ==
uint64_t it could be possible,
overflow check here doesn't harm in any way and saves headache of debugging
when overflow happens.
>
> Regards,
> Bharata.
>
