Re: [Qemu-devel] [PATCH v4 01/18] migration: dump vmstate info as a json file for static analysis

[Amit Shah]

On (Wed) 18 Jun 2014 [12:24:25], Juan Quintela wrote:
> Amit Shah <address@hidden> wrote:
> > This commit adds a new command, '-dump-vmstate', that takes a filename
> > as a parameter.  When executed, QEMU will dump the vmstate information
> > for the machine type it's invoked with to the file, and quit.
> >
> > The JSON-format output can then be used to compare the vmstate info for
> > different QEMU versions, specifically to test whether live migration
> > would break due to changes in the vmstate data.
> >
> > A Python script that compares the output of such JSON dumps is included
> > in the following commit.
> >
> > Signed-off-by: Amit Shah <address@hidden>
> 
> 
> 
> > +static void dump_vmstate_vmsd(FILE *out_file,
> > +                              const VMStateDescription *vmsd, int indent,
> > +                              bool is_subsection)
> > +{
> > +    if (is_subsection) {
> > +        fprintf(out_file, "%*s{\n", indent, "");
> > +    } else {
> > +        fprintf(out_file, "%*s\"%s\": {\n", indent, "", "Description");
> > +    }
> > +    indent += 2;
> > +    fprintf(out_file, "%*s\"name\": \"%s\",\n", indent, "", vmsd->name);
> > +    fprintf(out_file, "%*s\"version_id\": %d,\n", indent, "",
> > +            vmsd->version_id);
> > +    fprintf(out_file, "%*s\"minimum_version_id\": %d", indent, "",
> > +            vmsd->minimum_version_id);
> > +    if (vmsd->fields != NULL) {
> > +        const VMStateField *field = vmsd->fields;
> > +        bool first;
> > +
> > +        fprintf(out_file, ",\n%*s\"Fields\": [\n", indent, "");
> 
> Remove last "\n" (*)
> 
> 
> > +        first = true;
> 
> first can go now
> 
> > +        while (field->name != NULL) {
> > +            if (field->flags & VMS_MUST_EXIST) {
> > +                /* Ignore VMSTATE_VALIDATE bits; these don't get migrated 
> > */
> > +                field++;
> > +                continue;
> > +            }
> > +            if (!first) {
> > +                fprintf(out_file, ",\n");
> 
> You can print always \n now, right?

There's also a , there...

This sequence was added recently (v2 onwards) for the ignoring of the
VMS_MUST_EXIST stuff.

> Same for the other places?  Or I am missing something.
> 
> I will even go that itwould be better to just left the \n on the (*),
> and just add this \n at the end of writing a subsection.
> 
> 
> > +    fprintf(out_file, "\n%*s}", indent - 2, "");
> 
> And you remove it from here.

I tried several things with the \n; the current setting is the best I
found.

Of course, this is just pretty-printing, so I don't actually remember
all the details but I can look it up my git tree...

> <full disclosure>
> Yes, I have always hated pretty-printers
> </full-disclosure>

Oh, me too!

> Rest of this, I fully agree.
> 
> Later, Juan.

                Amit