Re: [Qemu-block] ping [PATCH v14] block/raw-posix.c: Make physical devices usable in QEMU under Mac OS X host

[Kevin Wolf]

Am 02.03.2016 um 04:32 hat Programmingkid geschrieben:
> 
> On Mar 1, 2016, at 10:16 AM, Kevin Wolf wrote:
> 
> > Am 29.02.2016 um 16:17 hat Programmingkid geschrieben:
> >> I do think this patch is ready to be added to QEMU. I have listened to 
> >> what you said and implemented your changes. 
> >> 
> >> https://patchwork.ozlabs.org/patch/579325/
> >> 
> >> Mac OS X can be picky when it comes to allowing the user
> >> to use physical devices in QEMU. Most mounted volumes
> >> appear to be off limits to QEMU. If an issue is detected,
> >> a message is displayed showing the user how to unmount a
> >> volume. Now QEMU uses both CD and DVD media.
> >> 
> >> Signed-off-by: John Arbuckle <address@hidden>
> >> 
> >> ---
> >> Changed filename variable to const char * type.
> >> Removed snprintf call for filename variable.
> >> filename is set to bsd_path if using a physical device that isn't a DVD or 
> >> CD.
> > 
> >> @@ -2112,33 +2166,57 @@ static int hdev_open(BlockDriverState *bs, QDict 
> >> *options, int flags,
> >> 
> >> #if defined(__APPLE__) && defined(__MACH__)
> >>     const char *filename = qdict_get_str(options, "filename");
> >> +    char bsd_path[MAXPATHLEN] = "";
> >> +    bool error_occurred = false;
> >> +    
> > 
> > This line adds trailing whitespace.
> > 
> >> @@ -2147,7 +2225,16 @@ static int hdev_open(BlockDriverState *bs, QDict 
> >> *options, int flags,
> >>         if (local_err) {
> >>             error_propagate(errp, local_err);
> >>         }
> >> -        return ret;
> >> +#if defined(__APPLE__) && defined(__MACH__)
> >> +        if (*bsd_path) {
> >> +            filename = bsd_path;
> >> +        }
> >> +        /* if a physical device experienced an error while being opened */
> >> +        if (strncmp(filename, "/dev/", 5) == 0) {
> >> +            print_unmounting_directions(filename);
> >> +            return -1;
> > 
> > Please use a negative errno number instead of -1.
> 
> Is this ok:
>       return -EPERM;
> 
> According to http://man7.org/linux/man-pages/man3/errno.3.html, it means 
> "operation not permitted".

Well, to be honest I don't understand why there is a different error
code here to begin with. Maybe when you add the "return ret" back after
the #endif you can just leave out this line and return the real error
code this way.

If for some reason (that I fail to understand) ret doesn't contain an
appropriate error code in this case, though, you can return a constant.
If it's related to permissions, -EPERM is okay, otherwise it's probably
not. I don't see a connection between paths starting with /dev/ and
there being a permission problem.

> Did you want this -1 changed also?
> +hdev_open_Mac_error:
> +        g_free(mediaType);
> +        if (mediaIterator) {
> +            IOObjectRelease(mediaIterator);
> +        }
> +        if (error_occurred) {
> +            return -1;
> +        }

Yes, please, I missed that one. We don't have a valid ret here, so maybe
-ENOENT would be the closest one?

> > But more importantly: What happened with the return that you removed
> > above? Even in the non-Apple case, we don't return an error now, but
> > continue in this function. This looks certainly wrong. Did you intend to
> > move it to below the #ifdef block?
> 
> Good catch. I will place it right below the #endif /* defined(__APPLE__) && 
> defined(__MACH__) */. 

Thanks.

Kevin