1SG_WRITE_LONG(8) SG3_UTILS SG_WRITE_LONG(8)
2
6 sg_write_long - send the SCSI WRITE LONG command
7
9 sg_write_long [--16] [--cor_dis] [--help] [--in=IF] [--lba=LBA]
10 [--pblock] [--verbose] [--version] [--wr_uncor] [--xfer_len=BTL] DEVICE
11
13 Send the SCSI WRITE LONG (10 or 16 byte) command to DEVICE. The buffer
14 to be written to the DEVICE is filled with 0xff bytes or read from the
15 IF file. This buffer includes the logical data (e.g. 512 bytes) and the
16 ECC bytes.
17 This utility can be used to generate a MEDIUM ERROR at a specific logi‐
19 cal block address. This can be useful for testing error handling. Prior
20 to such a test, the sg_dd utility could be used to copy the original
21 contents of the logical block address to some safe location. After the
22 test the sg_dd utility could be used to write back the original con‐
23 tents of the logical block address. An alternate strategy would be to
24 read the "long" contents of the logical block address with sg_read_long
25 utility prior to testing and restore it with this utility after test‐
26 ing.
27 Take care: If recoverable errors are being injected (e.g. only one or a
29 few bits changed so that the ECC is able to correct the data) then care
30 should be taken with the settings in the "read write error recovery"
31 mode page. Specifically if the ARRE (for reads) and/or AWRE (for
32 writes) are set then recovered errors will cause the lba to be reas‐
33 signed (and the old location to be added to the grown defect list
34 (PLIST)). This is not easily reversed and uses (one of the finite num‐
35 ber of) the spare sectors set aside for this purpose. If in doubt it is
36 probably safest to clear the ARRE and AWRE bits. These bits can be
37 checked and modified with the sdparm utility. For example: "sdparm -c
38 AWRE,ARRE /dev/sda" will clear the bits until the disk is power cycled.
39
41 Arguments to long options are mandatory for short options as well.
42 -S, --16
44 send a SCSI WRITE LONG (16) command to DEVICE. The default
45 action (in the absence of this option) is to send a SCSI WRITE
46 LONG (10) command.
47 -c, --cor_dis
49 sets the correction disabled (i.e 'COR_DIS') bit. This inhibits
50 various other mechanisms such as automatic block reallocation,
51 error recovery and various informational exception conditions
52 being triggered. This bit is relatively new in SBC-3 .
53 -h, --help
55 output the usage message then exit.
56 -i, --in=IF
58 read data (binary) from file named IF and use it for the SCSI
59 WRITE LONG command. If IF is "-" then stdin is read. If this
60 option is not given then 0xff bytes are used as fill.
61 -l, --lba=LBA
63 where LBA is the logical block address of the sector to over‐
64 write. Defaults to lba 0 which is a dangerous block to over‐
65 write on a disk that is in use. Assumed to be in decimal unless
66 prefixed with '0x' or has a trailing 'h'. If LBA is larger than
67 can fit in 32 bits then the --16 option should be used.
68 -p, --pblock
70 sets the physical block (i.e 'PBLOCK') bit. This instructs
71 DEVICE to use the given data (unless --wr_uncor is also given)
72 to write to the physical block specified by LBA. The default
73 action is to write to the logical block corresponding to the
74 given lba. This bit is relatively new in SBC-3 .
75 -v, --verbose
77 increase the degree of verbosity (debug messages).
78 -V, --version
80 output version string then exit.
81 -w, --wr_uncor
83 sets the "write uncorrected" (i.e 'WR_UNCOR') bit. This
84 instructs the DEVICE to flag the given lba (or the physical
85 block that contains it if --pblock is also given) as having an
86 unrecoverable error associated with it. Note: no data is trans‐
87 ferred to DEVICE, other than the command (i.e. the cdb). In the
88 absence of this option, the default action is to use the pro‐
89 vided data or 0xff bytes (--xfer_len=BTL in length) and write it
90 to DEVICE. This bit is relatively new in SBC-3 .
91 -x, --xfer_len=BTL
93 where BTL is the byte transfer length (default to 520). If the
94 given value (or the default) does not match the "long" block
95 size of the device, nothing is written to DEVICE and the appro‐
96 priate xfer_len value may be deduced from the error response
97 which is printed (to stderr).
98
100 Various numeric arguments (e.g. LBA) may include multiplicative suf‐
101 fixes or be given in hexadecimal. See the "NUMERIC ARGUMENTS" section
102 in the sg3_utils(8) man page.
103 The 10 byte SCSI WRITE LONG command limits the logical block address to
105 a 32 bit quantity. For larger LBAs use the --16 option for the SCSI
106 WRITE LONG (16) command.
107
109 This section outlines setting up a block with corrupted data, checking
110 the error condition, then restoring useful contents to that sector.
111 First, if the data in a sector is important, save it with the
113 sg_read_long utility:
114 sg_read_long --lba=0x1234 --out=0x1234_1.img -x BTL /dev/sda
116 This utility may need to be executed several time in order to determine
118 what the correct value for BTL is. Next use this utility to "corrupt"
119 that sector. That might be done with:
120 sg_write_long --lba=0x1234 -x BTL /dev/sda
122 This will write a sector (and ECC data) of 0xff bytes. Some disks may
124 reject this (at least one of the author's does). Another approach is to
125 copy the 0x1234_1.img file (to 0x1234_2.img in this example) and change
126 some values with a hex editor. Then write the changed image with:
127 sg_write_long --lba=0x1234 --in=0x1234_2.img -x BTL /dev/sda
129 Yet another approach is to use the --wr_uncor option, if supported:
131 sg_write_long --lba=0x1234 --wr_uncor /dev/sda
133 Next we use the sg_dd utility to check that the sector is corrupted.
135 Here is an example:
136 sg_dd if=/dev/sda blk_sgio=1 skip=0x1234 of=. bs=512 count=1 ver‐
138 bose=4
139 Notice that the "blk_sgio=1" option is given. This is to make sure that
141 the sector is read (and no others) and the error is fully reported.
142 The "blk_sgio=1" option causes the SG_IO ioctl to be used by sg_dd
143 rather than the block subsystem.
144 Finally we should restore sector 0x1234 to a non-corrupted state. A
146 sector full of zeroes could be written with:
147 sg_dd if=/dev/zero of=/dev/sda blk_sgio=1 seek=0x1234 bs=512 count=1
149 This will result in a sector (block) with 512 bytes of 0x0 without a
151 MEDIUM ERROR since the ECC and associated data will be regenerated and
152 thus well formed. The 'blk_sgio=1' option is even more important in
153 this case as it may stop the block subsystem doing a read before write
154 (since the read will most likely fail). Another approach is to write
155 back the original contents:
156 sg_write_long --lba=0x1234 --in=0x1234_1.img -x BTL /dev/sda
158
160 The exit status of sg_write_long is 0 when it is successful. Otherwise
161 see the sg3_utils(8) man page.
162
164 Written by Saeed Bishara. Further work by Douglas Gilbert.
165
167 Report bugs to <dgilbert at interlog dot com>.
168
170 Copyright © 2004-2009 Douglas Gilbert
171 This software is distributed under the GPL version 2. There is NO war‐
172 ranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PUR‐
173 POSE.
174
176 sg_read_long, sg_dd (both in sg3_utils), sdparm(sdparm)
177sg3_utils-1.27 March 2009 SG_WRITE_LONG(8)