On 2019/1/9 12:42, Jaegeuk Kim wrote:
> On 01/09, Chao Yu wrote:
>> On 2019/1/9 7:46, Jaegeuk Kim wrote:
>>> On 12/10, Chao Yu wrote:
>>>> stat() can fail due to a lot of reasons, let f2fs_dev_is_umounted()
>>>> detect such error and handle it correctly.
>>>>
>>>> Signed-off-by: Chao Yu <yuch...@huawei.com>
>>>> ---
>>>> lib/libf2fs.c | 9 ++++++++-
>>>> 1 file changed, 8 insertions(+), 1 deletion(-)
>>>>
>>>> diff --git a/lib/libf2fs.c b/lib/libf2fs.c
>>>> index c692bf2da635..6a1040feff32 100644
>>>> --- a/lib/libf2fs.c
>>>> +++ b/lib/libf2fs.c
>>>> @@ -713,7 +713,14 @@ int f2fs_dev_is_umounted(char *path)
>>>> st_buf = malloc(sizeof(struct stat));
>>>> ASSERT(st_buf);
>>>>
>>>> - if (stat(path, st_buf) == 0 && S_ISBLK(st_buf->st_mode)) {
>>>> + ret = stat(path, st_buf);
>>>
>>> We have to allow creating non-existing image file to build userdata.img.
>>> Can we
>>> check its errno?
>>
>> You mean sload flow? so how about?
>
> mkfs.f2fs on non-existing image file?
That's not allowed as e2fsprogs does?
mkfs.ext4 /home/image/img
mke2fs 1.44.4 (18-Aug-2018)
The file /home/image/img does not exist and no size was specified.
>
>>
>> if (ret) {
>> if (errno == ENOENT && c.func == SLOAD)
>> return 0;
>> MSG(0, "\tError: stat %s failed!\n", path);
>> free(st_buf);
>> return -1;
>> }
>>
>> Thanks,
>>
>>>
>>>> + if (ret) {
>>>> + MSG(0, "\tError: stat %s failed!\n", path);
>>>> + free(st_buf);
>>>> + return -1;
>>>> + }
>>>> +
>>>> + if (S_ISBLK(st_buf->st_mode)) {
>>>> int fd = open(path, O_RDONLY | O_EXCL);
>>>>
>>>> if (fd >= 0) {
>>>> --
>>>> 2.18.0.rc1
>>>
>>> .
>>>
>
> .
>
_______________________________________________
Linux-f2fs-devel mailing list
Linux-f2fs-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel
