Re: [Qemu-devel] [PULL 54/61] blockdev: Remove IF_* check for read-only blockdev_init

[Stefan Weil <sw@weilnetz.de>]

Am 15.10.2013 17:59, schrieb Kevin Wolf:
> Am 15.10.2013 um 17:53 hat Stefan Weil geschrieben:
>> Am 11.10.2013 17:05, schrieb Kevin Wolf:
>>> IF_NONE allows read-only, which makes forbidding it in this place
>>> for other types pretty much pointless.
>>>
>>> Instead, make sure that all devices for which the check would have
>>> errored out check in their init function that they don't get a read-only
>>> BlockDriverState. This catches even cases where IF_NONE and -device is
>>> used.
>>>
>>> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
>>> Reviewed-by: Eric Blake <eblake@redhat.com>
>>> ---
>>>
>> This patch breaks current QEMU (SIGSEGV with ARM in several test scenarios):
>>
>> $ git bisect bad
>> 4f8a066b5fc254eeaabbbde56ba4f5b29cc68fdf is the first bad commit
>> commit 4f8a066b5fc254eeaabbbde56ba4f5b29cc68fdf
>> Author: Kevin Wolf <kwolf@redhat.com>
>> Date:   Fri Sep 13 15:51:47 2013 +0200
>>
>>     blockdev: Remove IF_* check for read-only blockdev_init
>> [...]
>>
>> See the gdb protocol below for more details (Linux x86_64 host, default
>> configuration).
>>
>> I got a bug report from a Windows user, but the crash is not OS specific.
> Does this fix the segfault?
>
> diff --git a/hw/sd/sd.c b/hw/sd/sd.c
> index 7380f06..4502ad1 100644
> --- a/hw/sd/sd.c
> +++ b/hw/sd/sd.c
> @@ -494,7 +494,7 @@ SDState *sd_init(BlockDriverState *bs, bool is_spi)
>  {
>      SDState *sd;
>  
> -    if (bdrv_is_read_only(bs)) {
> +    if (bs && bdrv_is_read_only(bs)) {
>          fprintf(stderr, "sd_init: Cannot use read-only drive\n");
>          return NULL;
>      }

Yes.