1. Patchew
  2. QEMU
  3. [PULL 00/27] ppc-for-20250928 queue
  4. View patch

[PULL 19/27] ppc/spapr: init lrdr-capapcity phys with ram size if maxmem not provided

Harsh Prateek Bora posted 27 patches 1 month, 2 weeks ago
Maintainers: Nicholas Piggin <npiggin@gmail.com>, Aditya Gupta <adityag@linux.ibm.com>, Glenn Miles <milesg@linux.ibm.com>, Gautam Menghani <gautam@linux.ibm.com>, Harsh Prateek Bora <harshpb@linux.ibm.com>, Paolo Bonzini <pbonzini@redhat.com>, Chinmay Rath <rathc@linux.ibm.com>
[PULL 19/27] ppc/spapr: init lrdr-capapcity phys with ram size if maxmem not provided
Posted by Harsh Prateek Bora 1 month, 2 weeks ago 
lrdr-capacity contains phys field which communicates the maximum address
in bytes and therefore, the most memory that can be allocated to this
partition. This is usually populated when maxmem is provided alongwith
memory size on qemu command line. However since maxmem is an optional
param, this leads to bits being set to 0 in absence of maxmem param.
Fix this by initializing the respective bits as per total mem size in
such case.

Reported-by: Gaurav Batra <gbatra@us.ibm.com>
Tested-by: David Christensen <drc@linux.ibm.com>
Signed-off-by: Harsh Prateek Bora <harshpb@linux.ibm.com>
Reviewed-by: Shivaprasad G Bhat <sbhat@linux.ibm.com>
Link: https://lore.kernel.org/r/20250506042903.76250-1-harshpb@linux.ibm.com
Message-ID: <20250506042903.76250-1-harshpb@linux.ibm.com>
---
 hw/ppc/spapr.c | 11 +++++++----
 1 file changed, 7 insertions(+), 4 deletions(-)

diff --git a/hw/ppc/spapr.c b/hw/ppc/spapr.c
index eb22333404..82fb23beaa 100644
--- a/hw/ppc/spapr.c
+++ b/hw/ppc/spapr.c
@@ -907,6 +907,7 @@ static void spapr_dt_rtas(SpaprMachineState *spapr, void *fdt)
     int rtas;
     GString *hypertas = g_string_sized_new(256);
     GString *qemu_hypertas = g_string_sized_new(256);
+    uint64_t max_device_addr = 0;
     uint32_t lrdr_capacity[] = {
         0,
         0,
@@ -917,13 +918,15 @@ static void spapr_dt_rtas(SpaprMachineState *spapr, void *fdt)
 
     /* Do we have device memory? */
     if (MACHINE(spapr)->device_memory) {
-        uint64_t max_device_addr = MACHINE(spapr)->device_memory->base +
+        max_device_addr = MACHINE(spapr)->device_memory->base +
             memory_region_size(&MACHINE(spapr)->device_memory->mr);
-
-        lrdr_capacity[0] = cpu_to_be32(max_device_addr >> 32);
-        lrdr_capacity[1] = cpu_to_be32(max_device_addr & 0xffffffff);
+    } else if (ms->ram_size == ms->maxram_size) {
+        max_device_addr = ms->ram_size;
     }
 
+    lrdr_capacity[0] = cpu_to_be32(max_device_addr >> 32);
+    lrdr_capacity[1] = cpu_to_be32(max_device_addr & 0xffffffff);
+
     _FDT(rtas = fdt_add_subnode(fdt, 0, "rtas"));
 
     /* hypertas */
-- 
2.43.5
Re: [PULL 19/27] ppc/spapr: init lrdr-capapcity phys with ram size if maxmem not provided
Posted by Michael Tokarev 1 month, 2 weeks ago 
On 9/28/25 22:26, Harsh Prateek Bora wrote:
> lrdr-capacity contains phys field which communicates the maximum address
> in bytes and therefore, the most memory that can be allocated to this
> partition. This is usually populated when maxmem is provided alongwith
> memory size on qemu command line. However since maxmem is an optional
> param, this leads to bits being set to 0 in absence of maxmem param.
> Fix this by initializing the respective bits as per total mem size in
> such case.
> 
> Reported-by: Gaurav Batra <gbatra@us.ibm.com>
> Tested-by: David Christensen <drc@linux.ibm.com>
> Signed-off-by: Harsh Prateek Bora <harshpb@linux.ibm.com>
> Reviewed-by: Shivaprasad G Bhat <sbhat@linux.ibm.com>
> Link: https://lore.kernel.org/r/20250506042903.76250-1-harshpb@linux.ibm.com
> Message-ID: <20250506042903.76250-1-harshpb@linux.ibm.com>

This feels like a qemu-stable matherial (for 10.0 & 10.1 series).
Please let me know if it isn't.

Thanks,

/mjt

> ---
>   hw/ppc/spapr.c | 11 +++++++----
>   1 file changed, 7 insertions(+), 4 deletions(-)
> 
> diff --git a/hw/ppc/spapr.c b/hw/ppc/spapr.c
> index eb22333404..82fb23beaa 100644
> --- a/hw/ppc/spapr.c
> +++ b/hw/ppc/spapr.c
> @@ -907,6 +907,7 @@ static void spapr_dt_rtas(SpaprMachineState *spapr, void *fdt)
>       int rtas;
>       GString *hypertas = g_string_sized_new(256);
>       GString *qemu_hypertas = g_string_sized_new(256);
> +    uint64_t max_device_addr = 0;
>       uint32_t lrdr_capacity[] = {
>           0,
>           0,
> @@ -917,13 +918,15 @@ static void spapr_dt_rtas(SpaprMachineState *spapr, void *fdt)
>   
>       /* Do we have device memory? */
>       if (MACHINE(spapr)->device_memory) {
> -        uint64_t max_device_addr = MACHINE(spapr)->device_memory->base +
> +        max_device_addr = MACHINE(spapr)->device_memory->base +
>               memory_region_size(&MACHINE(spapr)->device_memory->mr);
> -
> -        lrdr_capacity[0] = cpu_to_be32(max_device_addr >> 32);
> -        lrdr_capacity[1] = cpu_to_be32(max_device_addr & 0xffffffff);
> +    } else if (ms->ram_size == ms->maxram_size) {
> +        max_device_addr = ms->ram_size;
>       }
>   
> +    lrdr_capacity[0] = cpu_to_be32(max_device_addr >> 32);
> +    lrdr_capacity[1] = cpu_to_be32(max_device_addr & 0xffffffff);
> +
>       _FDT(rtas = fdt_add_subnode(fdt, 0, "rtas"));
>   
>       /* hypertas */
Re: [PULL 19/27] ppc/spapr: init lrdr-capapcity phys with ram size if maxmem not provided
Posted by Harsh Prateek Bora 1 month, 2 weeks ago 

On 9/30/25 01:06, Michael Tokarev wrote:
> On 9/28/25 22:26, Harsh Prateek Bora wrote:
>> lrdr-capacity contains phys field which communicates the maximum address
>> in bytes and therefore, the most memory that can be allocated to this
>> partition. This is usually populated when maxmem is provided alongwith
>> memory size on qemu command line. However since maxmem is an optional
>> param, this leads to bits being set to 0 in absence of maxmem param.
>> Fix this by initializing the respective bits as per total mem size in
>> such case.
>>
>> Reported-by: Gaurav Batra <gbatra@us.ibm.com>
>> Tested-by: David Christensen <drc@linux.ibm.com>
>> Signed-off-by: Harsh Prateek Bora <harshpb@linux.ibm.com>
>> Reviewed-by: Shivaprasad G Bhat <sbhat@linux.ibm.com>
>> Link: 
>> https://lore.kernel.org/r/20250506042903.76250-1-harshpb@linux.ibm.com
>> Message-ID: <20250506042903.76250-1-harshpb@linux.ibm.com>
> 
> This feels like a qemu-stable matherial (for 10.0 & 10.1 series).
> Please let me know if it isn't.

Yes, it can be picked for stable releases as well. Thanks.

regards,
Harsh

> 
> Thanks,
> 
> /mjt
> 
>> ---
>>   hw/ppc/spapr.c | 11 +++++++----
>>   1 file changed, 7 insertions(+), 4 deletions(-)
>>
>> diff --git a/hw/ppc/spapr.c b/hw/ppc/spapr.c
>> index eb22333404..82fb23beaa 100644
>> --- a/hw/ppc/spapr.c
>> +++ b/hw/ppc/spapr.c
>> @@ -907,6 +907,7 @@ static void spapr_dt_rtas(SpaprMachineState 
>> *spapr, void *fdt)
>>       int rtas;
>>       GString *hypertas = g_string_sized_new(256);
>>       GString *qemu_hypertas = g_string_sized_new(256);
>> +    uint64_t max_device_addr = 0;
>>       uint32_t lrdr_capacity[] = {
>>           0,
>>           0,
>> @@ -917,13 +918,15 @@ static void spapr_dt_rtas(SpaprMachineState 
>> *spapr, void *fdt)
>>       /* Do we have device memory? */
>>       if (MACHINE(spapr)->device_memory) {
>> -        uint64_t max_device_addr = MACHINE(spapr)->device_memory->base +
>> +        max_device_addr = MACHINE(spapr)->device_memory->base +
>>               memory_region_size(&MACHINE(spapr)->device_memory->mr);
>> -
>> -        lrdr_capacity[0] = cpu_to_be32(max_device_addr >> 32);
>> -        lrdr_capacity[1] = cpu_to_be32(max_device_addr & 0xffffffff);
>> +    } else if (ms->ram_size == ms->maxram_size) {
>> +        max_device_addr = ms->ram_size;
>>       }
>> +    lrdr_capacity[0] = cpu_to_be32(max_device_addr >> 32);
>> +    lrdr_capacity[1] = cpu_to_be32(max_device_addr & 0xffffffff);
>> +
>>       _FDT(rtas = fdt_add_subnode(fdt, 0, "rtas"));
>>       /* hypertas */
> 
>

Patchew 2.1-devel-b67e4c2

© 2016 - 2025 Red Hat, Inc.