ZFS data error without reasons

[kevin]

Hi,
Will any changes cause zfs data error?I find my disk data error without 
any reasons(shutdown or reboot  normally).disk was bought 
yesterday.sometimes it can be fixed with a zpool scrub.but mostly zpool 
scrub will return more errors.Even i restore all zpool from my 
backup,without 5 mins,zpool status shows data error and many checksum 
errors.

Now i build kernel with following patch 1 hour ago . I get a kernel 
panic when zfs send snapshot from usb device to hard disk.

--- //depot/user/jhb/lock/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_dir.c
+++ /Users/jhb/work/p4/lock/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_dir.c

@@ -940,6 +940,7 @@
        /* NB: we already did dmu_tx_wait() if necessary */
        goto top;
    }
+    VOP_UNLOCK(*xvpp, 0);

    return (error);
 }


db>bt
Tracing pid 12 tid 100036 td 0xffffff0004083000
pmp_pvh_free() at pmp_pvh_free+0xa
pmp_remove_entry() at pmp_remove_entry+0x1b
pmp_enter() at pmp_enter+0x593
vm_fault() at vm_fault+0x1470
trap() at trap + 0x1dc
calltrap() at calltrap+0x8
--- trap 0xc, rip =0xffffffff80531e56, rsp = 0xfffffffee6346ab0, rbp = 
0xfffffffee6346af0 ---
bcopy() at bcopy+0x16
ata_dmaunload() at ata_dmaunload+0x53
ata_ahci_end_transaction() at ata_ahci_end_transaction+0x9d
ata_interrupt() at ata_interrupt+0xc4
ata_generic_intr() at ata_generic_intr+0x2f
intr_event_execute_handlers() at intr_event_execute_handlers+0xfc
ithread_loop() at ithread_loop+0x8e
fork_exit() at fork_exit+0x118
fork_trampoline() at fork_trampoline+0xe
--- trap 0, rip = 0, rsp = 0xfffffffee6346d30,rbp = 0 ---

Thanks,
kevin