Re: FreeBSD FUSE calls truncate() on read-only files

2015-03-02 Thread Julian Elischer
On 2/27/15 7:32 AM, Benjamin Kaduk wrote: On Fri, 27 Feb 2015, Julian Elischer wrote: for example it caches information when it shouldn't, even from 'dynamic' file systems We had to change the code to disable it as our data is synthetic and might change between reads. fstat info is also cached

Re: FreeBSD FUSE calls truncate() on read-only files

2015-02-27 Thread Benjamin Kaduk
On Fri, 27 Feb 2015, Julian Elischer wrote: > for example it caches information when it shouldn't, even from 'dynamic' file > systems > We had to change the code to disable it as our data is synthetic and might > change between reads. > fstat info is also cached and confused our apps mightily. Y

Re: FreeBSD FUSE calls truncate() on read-only files

2015-02-26 Thread Julian Elischer
On 2/25/15 5:34 AM, Eggert, Lars wrote: Hi, this came up when trying to port tup (https://github.com/gittup/tup) to FreeBSD. Even though we are opening the file read-only with cat, FUSE calls truncate() on it, which modifies its mtime and this screws up tup. See https://github.com/gittup/tup/

FreeBSD FUSE calls truncate() on read-only files

2015-02-25 Thread Eggert, Lars
Hi, this came up when trying to port tup (https://github.com/gittup/tup) to FreeBSD. Even though we are opening the file read-only with cat, FUSE calls truncate() on it, which modifies its mtime and this screws up tup. See https://github.com/gittup/tup/issues/198 Anyone know why FreeBSD's FUSE