I've never seen a regular usb stick report a 4KB logical sector size. I can see an external hard drive connected via usb potentially reporting it because there's usually a usb/sata mux chip in those devices, and sometimes they even use a RAID mux in a degenerate configuration which might report a 4KB logical sector size (even if the HDD is 512 logical / 4KB physical). But I've never seen a normal usb stick report a 4KB logical sector size.
-Matt On Tue, May 2, 2017 at 9:38 AM, Tomohiro Kusumi <[email protected]> wrote: > You can see this with other fs too. > I saw this with ext2 (I mean ext2 on dfly) on usb stick. > > 2017-05-02 19:23 GMT+03:00 Matthew Dillon <[email protected]>: > > Interesting. The usb block device is actually reporting a sector size of > > 4KB. If you want to experiment a little, try changing the BBSIZE macro > in > > the ntfs module from 1024 to 4096 and recompiling it. The BBSIZE macro > is > > in /usr/src/sys/vfs/ntfs/ntfs.h (in DragonFlyBSD). If you are unsure > how to > > just recompile one module, then just recompile the whole kernel. > > > > I think the other block operations in ntfs are based on a block size > defined > > in the filesystem volume header so there's a chance that this might solve > > the issue. > > > > -Matt > > > > On Tue, May 2, 2017 at 7:36 AM, Rod Person <[email protected]> > wrote: > >> > >> Hi, > >> > >> I'm attempting to mount and external usb ntfs hard drive: > >> <Seagate Backup+ Desk 0406> Fixed Direct Access SCSI-6 device > >> > >> In FreeBSD I mount it using the fuse driver, but being it isn't > >> available in DragonFly, I'm just attempting with mount_ntfs. I get > >> Invalid argument not matter how I try and mount it. I noticed the error: > >> > >> dscheck(da8): b_bcount 1024 is not on a sector boundary (ssize 4096) > >> > >> showing up in the dmesg and once when attempting to mount in the > >> console before starting X. That error doesn't show if I try and mount > >> in xterm, only writes to dmesg. I assume the the sector boundary is why > >> I can't mount it, but is there any way around this? > >> > >> > >> atomizer-dfly# mount_ntfs /dev/da8 /mnt/usb > >> mount_ntfs: /dev/da8: Invalid argument > >> > >> also tried /dev/da8s1 same result > >> > >> atomizer-dfly# dmesg | grep dscheck > >> dscheck(da8): b_bcount 1024 is not on a sector boundary (ssize 4096) > >> dscheck(da8s1): b_bcount 1024 is not on a sector boundary (ssize 4096) > >> > >> > >> atomizer-dfly# fdisk da8s1 > >> ******* Working on device /dev/da8s1 ******* > >> parameters extracted from device are: > >> cylinders=45600 heads=255 sectors/track=63 (16065 blks/cyl) > >> > >> Figures below won't work with BIOS for partitions not in cyl 1 > >> parameters to be used for BIOS calculations are: > >> cylinders=45600 heads=255 sectors/track=63 (16065 blks/cyl) > >> > >> Media sector size is 4096 > >> Warning: BIOS sector numbering starts with sector 1 > >> Information from DOS bootblock is: > >> The data for partition 1 is: > >> sysid 79,(QNX 4.2 Tertiary) > >> start 1936269394, size 1836016416 (7171939 Meg), flag d > >> beg: cyl 335/ head 10/ sector 2; > >> end: cyl 327/ head 84/ sector 13 > >> The data for partition 2 is: > >> sysid 115,(unknown) > >> start 1917848077, size 544437093 (2126707 Meg), flag 70 > >> beg: cyl 371/ head 114/ sector 37; > >> end: cyl 256/ head 101/ sector 36 > >> The data for partition 3 is: > >> sysid 43,(unknown) > >> start 1818575915, size 544175136 (2125684 Meg), flag 43 > >> beg: cyl 364/ head 116/ sector 50; > >> end: cyl 372/ head 65/ sector 44 > >> The data for partition 4 is: > >> sysid 97,(Speed) > >> start 2844524554, size 54974 (214 Meg), flag 72 > >> beg: cyl 372/ head 101/ sector 51; > >> end: cyl 269/ head 114/ sector 52 > >> -- > >> Rod > >> > >> http://www.rodperson.com > >> > >> He who knows himself to be one way and pretends it is another way is a > >> thief who robs his own soul. > >> > >> The Mahabharata > >> Sakuntala 25 > > > > >
