Re: [Devel] [PATCH] fs/fuse kio: satisfy pure FALLOC_FL_KEEP_SIZE immediately

2018-12-25 Thread Kirill Tkhai
On 25.12.2018 12:45, Pavel Butsykin wrote: > > > On 25.12.2018 12:28, Kirill Tkhai wrote: >> On 24.12.2018 15:54, Pavel Butsykin wrote: >>> Fallocate without mutex lock can race with setattr size request, as a >>> result, >>> may be various problems, including incorrectly changed file size. At

Re: [Devel] [PATCH] fs/fuse kio: satisfy pure FALLOC_FL_KEEP_SIZE immediately

2018-12-25 Thread Kirill Tkhai
On 24.12.2018 15:54, Pavel Butsykin wrote: > Fallocate without mutex lock can race with setattr size request, as a result, > may be various problems, including incorrectly changed file size. At the same > time pure FALLOC_FL_KEEP_SIZE for vstorage is just nope, so we can immediately man 2

Re: [Devel] [PATCH] fs/fuse kio: satisfy pure FALLOC_FL_KEEP_SIZE immediately

2018-12-25 Thread Pavel Butsykin
On 25.12.2018 12:28, Kirill Tkhai wrote: > On 24.12.2018 15:54, Pavel Butsykin wrote: >> Fallocate without mutex lock can race with setattr size request, as a result, >> may be various problems, including incorrectly changed file size. At the same >> time pure FALLOC_FL_KEEP_SIZE for vstorage

Re: [Devel] [PATCH] fs/fuse kio: satisfy pure FALLOC_FL_KEEP_SIZE immediately

2018-12-25 Thread Kirill Tkhai
On 25.12.2018 13:00, Pavel Butsykin wrote: > > > On 25.12.2018 12:28, Kirill Tkhai wrote: >> On 24.12.2018 15:54, Pavel Butsykin wrote: >>> Fallocate without mutex lock can race with setattr size request, as a >>> result, >>> may be various problems, including incorrectly changed file size. At

Re: [Devel] [PATCH] fs/fuse kio: satisfy pure FALLOC_FL_KEEP_SIZE immediately

2018-12-25 Thread Pavel Butsykin
On 25.12.2018 12:28, Kirill Tkhai wrote: > On 24.12.2018 15:54, Pavel Butsykin wrote: >> Fallocate without mutex lock can race with setattr size request, as a result, >> may be various problems, including incorrectly changed file size. At the same >> time pure FALLOC_FL_KEEP_SIZE for vstorage

Re: [Devel] [PATCH] fs/fuse kio: deny unavailable ioctl's in kio

2018-12-25 Thread Pavel Butsykin
pls ignore this. On 25.12.2018 18:11, Pavel Butsykin wrote: > There are some ioctl's that were made to service the kio module, and they > should not be available to user app. > > Also there are two ioctls PCS_IOC_NOCSUMONREAD and PCS_IOC_NOWRITEDELAY that > must be handled in kio, otherwise they

[Devel] [PATCH] fs/fuse kio: deny unavailable ioctl's in kio

2018-12-25 Thread Pavel Butsykin
There are some ioctl's that were made to service the kio module, and they should not be available to user app. Also there are two ioctls PCS_IOC_NOCSUMONREAD and PCS_IOC_NOWRITEDELAY that must be handled in kio, otherwise they are meaningless. So let's return -EOPNOTSUPP unless we implement them