Re: [PATCH v2] tools/libs/guest: fix save and restore of pv domains after 32-bit de-support

[Andrew Cooper <andrew.cooper3@xxxxxxxxxx>]

On 07/06/2021 14:04, Jan Beulich wrote:
> On 07.06.2021 15:00, Juergen Gross wrote:
>> --- a/tools/libs/guest/xg_sr_common_x86_pv.c
>> +++ b/tools/libs/guest/xg_sr_common_x86_pv.c
>> @@ -149,27 +149,32 @@ int x86_pv_map_m2p(struct xc_sr_context *ctx)
>>  
>>      ctx->x86.pv.nr_m2p_frames = (M2P_CHUNK_SIZE >> PAGE_SHIFT) * m2p_chunks;
>>  
>> +    if ( ctx->x86.pv.levels == 3 )
>> +    {
> With this opening brace you no longer need ...
>
>>  #ifdef __i386__
>> -    /* 32 bit toolstacks automatically get the compat m2p */
>> -    ctx->x86.pv.compat_m2p_mfn0 = entries[0].mfn;
>> +        /* 32 bit toolstacks automatically get the compat m2p */
>> +        ctx->x86.pv.compat_m2p_mfn0 = entries[0].mfn;
>>  #else
>> -    /* 64 bit toolstacks need to ask Xen specially for it */
>> -    {
> ... this one, and hence you could avoid re-indenting ...
>
>> -        struct xen_machphys_mfn_list xmml = {
>> -            .max_extents = 1,
>> -            .extent_start = { &ctx->x86.pv.compat_m2p_mfn0 },
>> -        };
>> -
>> -        rc = do_memory_op(xch, XENMEM_machphys_compat_mfn_list,
>> -                          &xmml, sizeof(xmml));
>> -        if ( rc || xmml.nr_extents != 1 )
>> +        /* 64 bit toolstacks need to ask Xen specially for it */
>>          {
>> -            PERROR("Failed to get compat mfn list from Xen");
>> -            rc = -1;
>> -            goto err;
>> +            struct xen_machphys_mfn_list xmml = {
>> +                .max_extents = 1,
>> +                .extent_start = { &ctx->x86.pv.compat_m2p_mfn0 },
>> +            };
>> +
>> +            rc = do_memory_op(xch, XENMEM_machphys_compat_mfn_list,
>> +                              &xmml, sizeof(xmml));
>> +            if ( rc || xmml.nr_extents != 1 )
>> +            {
>> +                PERROR("Failed to get compat mfn list from Xen");
>> +                rc = -1;
>> +                goto err;
>> +            }
> ... all of this. Preferably with such reduced code churn,
> still/again:

I agree.  I can fix on commit, if you're happy with that.

Reviewed-by: Andrew Cooper <andrew.cooper3@xxxxxxxxxx>