Re: [Qemu-devel] [PATCH v5 1/3] ACPI: Add new ACPI structures and macros

["Michael S. Tsirkin" <mst@redhat.com>]

On Thu, Jul 13, 2017 at 05:33:30PM +0200, Laszlo Ersek wrote:
> On 07/13/17 14:00, gengdongjiu wrote:
> > Laszlo,
> >   Thank you for your review and comments.
> > 
> > 
> > On 2017/7/13 18:33, Laszlo Ersek wrote:
> >> On 07/12/17 04:08, Dongjiu Geng wrote:
> 
> [snip]
> 
> >>> --- a/include/qemu/uuid.h
> >>> +++ b/include/qemu/uuid.h
> >>> @@ -44,6 +44,17 @@ typedef struct {
> >>>  
> >>>  #define UUID_NONE "00000000-0000-0000-0000-000000000000"
> >>>  
> >>> +#define UUID_BE(a, b, c, d0, d1, d2, d3, d4, d5, d6, d7)        \
> >>> +{{{ ((a) >> 24) & 0xff, ((a) >> 16) & 0xff, ((a) >> 8) & 0xff, (a) & 0xff, \
> >>> +   ((b) >> 8) & 0xff, (b) & 0xff,                   \
> >>> +    ((c) >> 8) & 0xff, (c) & 0xff,                    \
> >>> +    (d0), (d1), (d2), (d3), (d4), (d5), (d6), (d7) } } }
> >>> +
> >>> +/* Platform Memory, this is from UEFI 2.6 N.2.2 Section Descriptor */
> >>> +#define UEFI_CPER_SEC_PLATFORM_MEM                   \
> >>> +    UUID_BE(0xA5BC1114, 0x6F64, 0x4EDE, 0xB8, 0x63, 0x3E, 0x83, \
> >>> +    0xED, 0x7C, 0x83, 0xB1)
> >>> +
> >>>  void qemu_uuid_generate(QemuUUID *out);
> >>>  
> >>>  int qemu_uuid_is_null(const QemuUUID *uu);
> >>>
> >>
> >> (e) I think the addition of UUID_BE should be split out to a separate
> >> patch; it adds a general facility. It should likely be the very first
> >> patch in the series.
> >   Ok.
> > 
> >>
> >> (f) While I think it is justified to have UUID_BE() in "qemu/uuid.h", I
> >> think UEFI_CPER_SEC_PLATFORM_MEM is too specific to have here.
> >>
> >> If UEFI_CPER_SEC_PLATFORM_MEM were *not* a standardized UUID, I would
> >> suggest moving it to the implementation ("include/hw/acpi/hest_ghes.h"
> >> -- which in turn should be moved to patch #2, see my remark (d)), *plus*
> >> I would suggest eliminating the new #include from "acpi-defs.h", see my
> >> remark (b).
> >   understand your idea.
> > 
> >>
> >> However, given that this UUID *is* standard, I suggest keeping the (b)
> >> #include as you currently propose, and to move the definition of
> >> UEFI_CPER_SEC_PLATFORM_MEM to "acpi-defs.h".
> >   I agree with you.
> > 
> >>
> >> I vaguely recall that Michael commented on this previously, but I don't
> >> remember what he said. Michael, are you OK with my suggestion?
> >    Laszlo, I pasted Michael's comments here, as shown below. Michael said the definition
> > should use build_append_int_noprefix to add data. but I think it may not good, becuase
> > the section "UEFI_CPER_SEC_PLATFORM_MEM" is runtime recorded as CPER, not a ACPI/HEST
> > table member, so it is not generated when system boot up.
> 
> I agree: the UUID in question is not placed into the ACPI payload /
> fw_cfg blobs, it is written into guest memory at runtime, into the
> firmware-allocated area, if and when there is a hardware error to report.
> 
> Thanks
> Laszlo

The main point is that wrapping it up in a macro with an
unreadable name is not really helpful when it's only
used in a single place.


> > On the other hand,UEFI_CPER_SEC_PLATFORM_MEM
> > definition is from UEFI spec 2.6, N.2.2 Section Descriptor: {0xA5BC1114, 0x6F64, 0x4EDE, {0xB8, 0x63, 0x3E, 0x83, 0xED, 0x7C, 0x83, 0xB1}}.
> > if use  build_append_int_noprefix to add, may confuse others.
> > 
> > -----------------------------------------------------------------
> > -----------------------------------------------------------------
> > There's no reason to define these messy one-time use macros.
> > They just make it hard to look things up in the spec.
> > 
> > 
> > You can use build_append_int_noprefix to add data of
> > any length in LE format.
> > -----------------------------------------------------------------
> > -----------------------------------------------------------------
> > 
> > Hi  Michael,
> >   what is your suggestion about it? do you agree with Laszlo?

My main point is that the macros do not seem helpful.