[Qemu-devel] [PATCH v6 0/2] spapr: rma and numa nodes fixes

[Qemu-devel] [PATCH v6 0/2] spapr: rma and numa nodes fixes

[Alexey Kardashevskiy]

This fixes a bug in patch#1 and mistype in patch#2, details are in
the commit messages.

Alexey Kardashevsksy (1):
  spapr: make sure RMA is in first mode of first memory node

Paul Mackerras (1):
  spapr: limit numa memory regions by ram size

 hw/ppc/spapr.c | 35 ++++++++++++++++++++++++++---------
 1 file changed, 26 insertions(+), 9 deletions(-)

-- 
1.8.4.rc4

[Qemu-devel] [PATCH v6 1/2] spapr: make sure RMA is in first mode of first memory node

[Alexey Kardashevskiy]

The SPAPR specification says that the RMA starts at the LPAR's logical
address 0 and is the first logical memory block reported in
the LPAR’s device tree.

So SLOF only maps the first block and that block needs to span
the full RMA.

This makes sure that the RMA area is where SLOF expects it.

Reviewed-by: Thomas Huth <thuth@linux.vnet.ibm.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
---
Changes:
v6:
* fixed a bug with spapr_reset_htab() which used to reset spapr->rma_size
to the entire RAM

v4:
* fixed a bug with preallocated RMA (thanks to Thomas Huth)

v3:
* removed unnecessary RMA fixup from spapr_populate_memory()

v2:
* changed as recommended by Alex Graf
---
 hw/ppc/spapr.c | 17 +++++++++++------
 1 file changed, 11 insertions(+), 6 deletions(-)

diff --git a/hw/ppc/spapr.c b/hw/ppc/spapr.c
index 7e53a5f..7426518 100644
--- a/hw/ppc/spapr.c
+++ b/hw/ppc/spapr.c
@@ -532,9 +532,6 @@ static int spapr_populate_memory(sPAPREnvironment *spapr, void *fdt)
 
     /* memory node(s) */
     node0_size = (nb_numa_nodes > 1) ? node_mem[0] : ram_size;
-    if (spapr->rma_size > node0_size) {
-        spapr->rma_size = node0_size;
-    }
 
     /* RMA */
     mem_reg_property[0] = 0;
@@ -688,7 +685,8 @@ static void spapr_reset_htab(sPAPREnvironment *spapr)
 
     /* Update the RMA size if necessary */
     if (spapr->vrma_adjust) {
-        spapr->rma_size = kvmppc_rma_size(ram_size, spapr->htab_shift);
+        hwaddr node0_size = (nb_numa_nodes > 1) ? node_mem[0] : ram_size;
+        spapr->rma_size = kvmppc_rma_size(node0_size, spapr->htab_shift);
     }
 }
 
@@ -1113,6 +1111,7 @@ static void ppc_spapr_init(QEMUMachineInitArgs *args)
     MemoryRegion *sysmem = get_system_memory();
     MemoryRegion *ram = g_new(MemoryRegion, 1);
     hwaddr rma_alloc_size;
+    hwaddr node0_size = (nb_numa_nodes > 1) ? node_mem[0] : ram_size;
     uint32_t initrd_base = 0;
     long kernel_size = 0, initrd_size = 0;
     long load_limit, rtas_limit, fw_size;
@@ -1134,10 +1133,10 @@ static void ppc_spapr_init(QEMUMachineInitArgs *args)
         exit(1);
     }
 
-    if (rma_alloc_size && (rma_alloc_size < ram_size)) {
+    if (rma_alloc_size && (rma_alloc_size < node0_size)) {
         spapr->rma_size = rma_alloc_size;
     } else {
-        spapr->rma_size = ram_size;
+        spapr->rma_size = node0_size;
 
         /* With KVM, we don't actually know whether KVM supports an
          * unbounded RMA (PR KVM) or is limited by the hash table size
@@ -1154,6 +1153,12 @@ static void ppc_spapr_init(QEMUMachineInitArgs *args)
         }
     }
 
+    if (spapr->rma_size > node0_size) {
+        fprintf(stderr, "Error: Numa node 0 has to span the RMA (%#08"HWADDR_PRIx")\n",
+                spapr->rma_size);
+        exit(1);
+    }
+
     /* We place the device tree and RTAS just below either the top of the RMA,
      * or just below 2GB, whichever is lowere, so that it can be
      * processed with 32-bit real mode code if necessary */
-- 
1.8.4.rc4

[Qemu-devel] [PATCH v6 2/2] spapr: limit numa memory regions by ram size

[Alexey Kardashevskiy]

From: Paul Mackerras <paulus@samba.org>

This makes sure that all NUMA memory blocks reside within RAM or
have zero length.

Reviewed-by: Thomas Huth <thuth@linux.vnet.ibm.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
---
Changes:
V6: fixed mistype

---

This is a bugfix for:
-m 500
-smp 8,sockets=2,cores=2,threads=2
-numa node,nodeid=0,cpus=0-3,mem=500
-numa node,nodeid=1,cpus=4-7,mem=500
---
 hw/ppc/spapr.c | 20 ++++++++++++++++----
 1 file changed, 16 insertions(+), 4 deletions(-)

diff --git a/hw/ppc/spapr.c b/hw/ppc/spapr.c
index 7426518..1239d80 100644
--- a/hw/ppc/spapr.c
+++ b/hw/ppc/spapr.c
@@ -526,12 +526,16 @@ static int spapr_populate_memory(sPAPREnvironment *spapr, void *fdt)
                                 cpu_to_be32(0x0), cpu_to_be32(0x0),
                                 cpu_to_be32(0x0)};
     char mem_name[32];
-    hwaddr node0_size, mem_start;
+    hwaddr node0_size, mem_start, node_size;
     uint64_t mem_reg_property[2];
     int i, off;
 
     /* memory node(s) */
-    node0_size = (nb_numa_nodes > 1) ? node_mem[0] : ram_size;
+    if (nb_numa_nodes > 1 && node_mem[0] < ram_size) {
+        node0_size = node_mem[0];
+    } else {
+        node0_size = ram_size;
+    }
 
     /* RMA */
     mem_reg_property[0] = 0;
@@ -563,7 +567,15 @@ static int spapr_populate_memory(sPAPREnvironment *spapr, void *fdt)
     mem_start = node0_size;
     for (i = 1; i < nb_numa_nodes; i++) {
         mem_reg_property[0] = cpu_to_be64(mem_start);
-        mem_reg_property[1] = cpu_to_be64(node_mem[i]);
+        if (mem_start >= ram_size) {
+            node_size = 0;
+        } else {
+            node_size = node_mem[i];
+            if (node_size > ram_size - mem_start) {
+                node_size = ram_size - mem_start;
+            }
+        }
+        mem_reg_property[1] = cpu_to_be64(node_size);
         associativity[3] = associativity[4] = cpu_to_be32(i);
         sprintf(mem_name, "memory@" TARGET_FMT_lx, mem_start);
         off = fdt_add_subnode(fdt, 0, mem_name);
@@ -573,7 +585,7 @@ static int spapr_populate_memory(sPAPREnvironment *spapr, void *fdt)
                           sizeof(mem_reg_property))));
         _FDT((fdt_setprop(fdt, off, "ibm,associativity", associativity,
                           sizeof(associativity))));
-        mem_start += node_mem[i];
+        mem_start += node_size;
     }
 
     return 0;
-- 
1.8.4.rc4

Re: [Qemu-devel] [PATCH v6 0/2] spapr: rma and numa nodes fixes

[Alexey Kardashevskiy]

On 11/25/2013 02:14 PM, Alexey Kardashevskiy wrote:
> This fixes a bug in patch#1 and mistype in patch#2, details are in
> the commit messages.
> 
> Alexey Kardashevsksy (1):
>   spapr: make sure RMA is in first mode of first memory node
> 
> Paul Mackerras (1):
>   spapr: limit numa memory regions by ram size
> 
>  hw/ppc/spapr.c | 35 ++++++++++++++++++++++++++---------
>  1 file changed, 26 insertions(+), 9 deletions(-)
> 

Alex, ping? It got ROB from Thomas :)


-- 
Alexey

Re: [Qemu-devel] [PATCH v6 0/2] spapr: rma and numa nodes fixes

[Alexey Kardashevskiy]

On 12/03/2013 02:45 PM, Alexey Kardashevskiy wrote:
> On 11/25/2013 02:14 PM, Alexey Kardashevskiy wrote:
>> This fixes a bug in patch#1 and mistype in patch#2, details are in
>> the commit messages.
>>
>> Alexey Kardashevsksy (1):
>>   spapr: make sure RMA is in first mode of first memory node
>>
>> Paul Mackerras (1):
>>   spapr: limit numa memory regions by ram size
>>
>>  hw/ppc/spapr.c | 35 ++++++++++++++++++++++++++---------
>>  1 file changed, 26 insertions(+), 9 deletions(-)
>>
> 
> Alex, ping? It got ROB from Thomas :)


Ping?


-- 
Alexey

Re: [Qemu-devel] [PATCH v6 0/2] spapr: rma and numa nodes fixes

[Alexander Graf]

On 10.12.2013, at 09:32, Alexey Kardashevskiy <aik@ozlabs.ru> wrote:

> On 12/03/2013 02:45 PM, Alexey Kardashevskiy wrote:
>> On 11/25/2013 02:14 PM, Alexey Kardashevskiy wrote:
>>> This fixes a bug in patch#1 and mistype in patch#2, details are in
>>> the commit messages.
>>> 
>>> Alexey Kardashevsksy (1):
>>>  spapr: make sure RMA is in first mode of first memory node
>>> 
>>> Paul Mackerras (1):
>>>  spapr: limit numa memory regions by ram size
>>> 
>>> hw/ppc/spapr.c | 35 ++++++++++++++++++++++++++---------
>>> 1 file changed, 26 insertions(+), 9 deletions(-)
>>> 
>> 
>> Alex, ping? It got ROB from Thomas :)
> 
> 
> Ping?

Thanks, applied to ppc-next.


Alex