|
From: | Richard Henderson |
Subject: | Re: [RFC PATCH 14/15] softmmu/cpus: Extract QMP command handlers to cpus-qmp.c |
Date: | Wed, 26 May 2021 14:47:35 -0700 |
User-agent: | Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101 Thunderbird/78.8.1 |
On 5/26/21 2:35 PM, Philippe Mathieu-Daudé wrote:
On 5/26/21 9:10 PM, Richard Henderson wrote:On 5/17/21 4:55 AM, Philippe Mathieu-Daudé wrote:qmp_memsave() and qmp_pmemsave() call cpu_memory_rw_debug() and cpu_physical_memory_read(), which are target specific prototypes.Is there any reason they should be?They use target_ulong. Should they use hwaddr instead?
cpu_physical_memory_* should use hwaddr.cpu_memory_rw_debug uses a virtual address, and so could probably use uint64_t, as per my comment vs patch 13 about not using hwaddr for virtual addresses.
r~
[Prev in Thread] | Current Thread | [Next in Thread] |