Re: [PATCH] hvf: remove unused but set variable

[Pierrick Bouvier <pierrick.bouvier@linaro.org>]

On 10/23/24 13:04, Pierrick Bouvier wrote:
> On 10/23/24 12:53, Alex Bennée wrote:
>> Pierrick Bouvier <pierrick.bouvier@linaro.org> writes:
>>
>>> fixes associated warning when building on MacOS.
>>>
>>> Signed-off-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
>>> ---
>>>    target/i386/hvf/x86_task.c | 10 +++++-----
>>>    1 file changed, 5 insertions(+), 5 deletions(-)
>>>
>>> diff --git a/target/i386/hvf/x86_task.c b/target/i386/hvf/x86_task.c
>>> index f09bfbdda5b..cdea2ea69d9 100644
>>> --- a/target/i386/hvf/x86_task.c
>>> +++ b/target/i386/hvf/x86_task.c
>>> @@ -122,7 +122,6 @@ void vmx_handle_task_switch(CPUState *cpu, x68_segment_selector tss_sel, int rea
>>>        load_regs(cpu);
>>>    
>>>        struct x86_segment_descriptor curr_tss_desc, next_tss_desc;
>>> -    int ret;
>>>        x68_segment_selector old_tss_sel = vmx_read_segment_selector(cpu, R_TR);
>>>        uint64_t old_tss_base = vmx_read_segment_base(cpu, R_TR);
>>>        uint32_t desc_limit;
>>> @@ -138,7 +137,7 @@ void vmx_handle_task_switch(CPUState *cpu, x68_segment_selector tss_sel, int rea
>>>        if (reason == TSR_IDT_GATE && gate_valid) {
>>>            int dpl;
>>>    
>>> -        ret = x86_read_call_gate(cpu, &task_gate_desc, gate);
>>> +        x86_read_call_gate(cpu, &task_gate_desc, gate);
>>
>> If we don't care and this is the only caller we fix up
>> x86_read_call_gate not to return a value. It looks like it fails safe
>> with an empty entry (but also the function x86_read_call_gate needs the
>> printf removing).
>>
> Yes, or maybe we should check if the read was a success, or else call
> VM_PANIC("x86_read_call_gate").
> Any advice from hvf maintainers?
> 

We didn't have any feedback yet.

Would that be possible for a maintainer to pull the current series as it is?
It's functionnally equivalent, and remove this warning to have a clean 
9.2 release on this platform.

>>>    
>>>            dpl = task_gate_desc.dpl;
>>>            x68_segment_selector cs = vmx_read_segment_selector(cpu, R_CS);
>>> @@ -167,11 +166,12 @@ void vmx_handle_task_switch(CPUState *cpu, x68_segment_selector tss_sel, int rea
>>>            x86_write_segment_descriptor(cpu, &next_tss_desc, tss_sel);
>>>        }
>>>    
>>> -    if (next_tss_desc.type & 8)
>>> -        ret = task_switch_32(cpu, tss_sel, old_tss_sel, old_tss_base, &next_tss_desc);
>>> -    else
>>> +    if (next_tss_desc.type & 8) {
>>> +        task_switch_32(cpu, tss_sel, old_tss_sel, old_tss_base, &next_tss_desc);
>>> +    } else {
>>>            //ret = task_switch_16(cpu, tss_sel, old_tss_sel, old_tss_base, &next_tss_desc);
>>>            VM_PANIC("task_switch_16");
>>> +    }
>>>    
>>>        macvm_set_cr0(cpu->accel->fd, rvmcs(cpu->accel->fd, VMCS_GUEST_CR0) |
>>>                                    CR0_TS_MASK);
>>