On 4/29/25 9:40 AM, Daniel Henrique Barboza wrote:
Joel,
I'll make these changes in this patch to be consistent with what we've
been discussing:
- change addr to hwaddr
- use HWADDR_PRIx instead of PRIx64
i.e. this diff:
diff --git a/hw/riscv/virt.c b/hw/riscv/virt.c
index 1eae84db15..0020d8f404 100644
--- a/hw/riscv/virt.c
+++ b/hw/riscv/virt.c
@@ -303,12 +303,13 @@ static void create_fdt_socket_cpus(RISCVVirtState *s, int
socket,
static void create_fdt_socket_memory(RISCVVirtState *s, int socket)
{
g_autofree char *mem_name = NULL;
- uint64_t addr, size;
+ hwaddr addr;
+ uint64_t size;
MachineState *ms = MACHINE(s);
addr = s->memmap[VIRT_DRAM].base + riscv_socket_mem_offset(ms, socket);
size = riscv_socket_mem_size(ms, socket);
- mem_name = g_strdup_printf("/memory@%"PRIx64, addr);
+ mem_name = g_strdup_printf("/memory@%"HWADDR_PRIx, addr);
qemu_fdt_add_subnode(ms->fdt, mem_name);
qemu_fdt_setprop_cells(ms->fdt, mem_name, "reg",
addr >> 32, addr, size >> 32, size);
I did this change and applied your "hw/riscv/virt: device tree reg cleanups"
series on top
of it, and there are no conflicts. No change needed in your side.
It seems I was wrong. The v2 will conflict with your patch 03. I think a rebase
from
your series can't be avoided ...
Daniel
Thanks,
Daniel
On 4/24/25 6:41 AM, Joel Stanley wrote:
On Wed, 23 Apr 2025 at 20:39, Daniel Henrique Barboza
<dbarb...@ventanamicro.com> wrote:
We can avoid the 'long' casts by using PRIx64 and HWADDR_PRIx on the fmt
strings for uint64_t and hwaddr types.
Signed-off-by: Daniel Henrique Barboza <dbarb...@ventanamicro.com>
---
hw/riscv/virt.c | 29 +++++++++++++++--------------
1 file changed, 15 insertions(+), 14 deletions(-)
diff --git a/hw/riscv/virt.c b/hw/riscv/virt.c
index 036a0a9bfb..075c035f25 100644
--- a/hw/riscv/virt.c
+++ b/hw/riscv/virt.c
@@ -309,7 +309,7 @@ static void create_fdt_socket_memory(RISCVVirtState *s, int
socket)
addr = s->memmap[VIRT_DRAM].base + riscv_socket_mem_offset(ms, socket);
size = riscv_socket_mem_size(ms, socket);
- mem_name = g_strdup_printf("/memory@%lx", (long)addr);
+ mem_name = g_strdup_printf("/memory@%"PRIx64, addr);
I wondered why this wasn't a HWADDR_PRIx.
addr (and NodeInfo::node_mem?) could be a hwaddr? That would make
everything more consistent.
Cheers,
Joel
