{ "mode": "man", "parameter": "sg_dd", "section": "8", "url": "https://www.chedong.com/phpMan.php/man/sg_dd/8/json", "generated": "2026-06-15T18:51:52Z", "synopsis": "sgdd [bs=BS] [conv=CONV] [count=COUNT] [ibs=BS] [if=IFILE] [iflag=FLAGS] [obs=BS] [of=OFILE]\n[oflag=FLAGS] [seek=SEEK] [skip=SKIP] [--help] [--verbose] [--version]\n[blksgio={0|1}] [bpt=BPT] [cdbsz={6|10|12|16}] [cdl=CDL] [coe={0|1|2|3}] [coelimit=CL]\n[dio={0|1}] [odir={0|1}] [of2=OFILE2] [retries=RETR] [sync={0|1}] [time={0|1}[,TO]] [ver‐\nbose=VERB] [--dry-run] [--progress] [--verify]", "sections": { "NAME": { "content": "sgdd - copy data to and from files and devices, especially SCSI devices\n", "subsections": [] }, "SYNOPSIS": { "content": "sgdd [bs=BS] [conv=CONV] [count=COUNT] [ibs=BS] [if=IFILE] [iflag=FLAGS] [obs=BS] [of=OFILE]\n[oflag=FLAGS] [seek=SEEK] [skip=SKIP] [--help] [--verbose] [--version]\n\n[blksgio={0|1}] [bpt=BPT] [cdbsz={6|10|12|16}] [cdl=CDL] [coe={0|1|2|3}] [coelimit=CL]\n[dio={0|1}] [odir={0|1}] [of2=OFILE2] [retries=RETR] [sync={0|1}] [time={0|1}[,TO]] [ver‐\nbose=VERB] [--dry-run] [--progress] [--verify]\n", "subsections": [] }, "DESCRIPTION": { "content": "Copy data to and from any files. Specialized for \"files\" that are Linux SCSI generic (sg) de‐\nvices, raw devices or other devices that support the SGIO ioctl (which are only found in the\nlk 2.6 series). Similar syntax and semantics to dd(1) command.\n\nThe first group in the synopsis above are \"standard\" Unix dd(1) operands. The second group\nare extra options added by this utility. Both groups are defined below.\n\nWhen the --verify option is given, then the read side is the same but the on the write side,\nthe WRITE SCSI command is replaced by the VERIFY SCSI command. If any VERIFY commands yields\na sense key of MISCOMPARE then the verify operation will stop. The --verify option can only\nbe used when OFILE is either a sg device or a block device with oflag=sgio also given. When\nthe --verify option is used, this utility works in a similar fashion to the Unix cmp(1) com‐\nmand.\n\nThis utility is only supported on Linux whereas most other utilities in the sg3utils package\nhave been ported to other operating systems. A utility called \"ddpt\" has similar syntax and\nfunctionality to sgdd. ddpt drops some Linux specific features while adding some other\ngeneric features. This allows ddpt to be ported to other operating systems.\n", "subsections": [] }, "OPTIONS": { "content": "blksgio={0|1}\nwhen set to 0, block devices (e.g. /dev/sda) are treated like normal files (i.e.\nread(2) and write(2) are used for IO). When set to 1, block devices are assumed to ac‐\ncept the SGIO ioctl and SCSI commands are issued for IO. This is only supported for\n2.6 series kernels. Note that ATAPI devices (e.g. cd/dvd players) use the SCSI command\nset but ATA disks do not (unless there is a protocol conversion as often occurs in the\nUSB mass storage class). If the input or output device is a block device partition\n(e.g. /dev/sda3) then setting this option causes the partition information to be ig‐\nnored (since access is directly to the underlying device). Default is 0. See the\n'sgio' flag.\n\nbpt=BPT\neach IO transaction will be made using BPT blocks (or less if near the end of the\ncopy). Default is 128 for logical block sizes less that 2048 bytes, otherwise the de‐\nfault is 32. So for bs=512 the reads and writes will each convey 64 KiB of data by de‐\nfault (less if near the end of the transfer or memory restrictions). When cd/dvd\ndrives are accessed, the logical block size is typically 2048 bytes and bpt defaults\nto 32 which again implies 64 KiB transfers. The block layer when the blksgio=1 option\nis used has relatively low upper limits for transfer sizes (compared to sg device\nnodes, see /sys/block//queue/maxsectorskb ).\n\nbs=BS where BS must be the logical block size of the physical device (if either the input or\noutput files are accessed via SCSI commands). Note that this differs from dd(1) which\npermits BS to be an integral multiple. Default is 512 which is usually correct for\ndisks but incorrect for cdroms (which normally have 2048 byte blocks). For this util‐\nity the maximum size of each individual IO operation is BS * BPT bytes.\n\ncdbsz={6|10|12|16}\nsize of SCSI READ and/or WRITE commands issued on sg device names (or block devices\nwhen 'iflag=sgio' and/or 'oflag=sgio' is given). Default is 10 byte SCSI command\nblocks (unless calculations indicate that a 4 byte block number may be exceeded or BPT\nis greater than 16 bits (65535), in which case it defaults to 16 byte SCSI commands).\n\ncdl=CDL\nallows setting of command duration limits. CDL is either a single value or two values\nseparated by a comma. If one value is given, it applies to both IFILE and OFILE (if\nthey are pass-through devices). If two values are given, the first applies to IFILE\nwhile the second applies to OFILE. The value may be from 0 to 7 where 0 is the default\nand means there are no command duration limits. Command duration limits are only sup‐\nported by 16 byte READ and WRITE commands (plus READ(32), WRITE(32) and the WRITE\nSCATTERED command, bit thay are used by this utility). If the cdbsz operand is not\ngiven and would have a value less than 16, then if CDL is greater than 0, the cdbsz is\nincreased to 16.\nCommand duration limits can be accesses and changed in the Command duration limit A\nand B mode pages, plus the Command duration limit T2A and T2B mode pages. The sdparm\nutility may be used to access and change these mode pages.\n\ncoe={0|1|2|3}\nset to 1 or more for continue on error ('coe'). Only applies to errors on sg devices\nor block devices with the 'sgio' flag set. Thus errors on other files will stop sgdd.\nDefault is 0 which implies stop on any error. See the 'coe' flag for more information.\n\ncoelimit=CL\nwhere CL is the maximum number of consecutive bad blocks stepped over (due to \"coe>0\")\non reads before the copy terminates. This only applies when IFILE is accessed via the\nSGIO ioctl. The default is 0 which is interpreted as no limit. This option is meant\nto stop the copy soon after unrecorded media is detected while still offering \"con‐\ntinue on error\" capability.\n\nconv=sparse\nsee the CONVERSIONS section below.\n\ncount=COUNT\ncopy COUNT blocks from IFILE to OFILE. Default is the minimum (of IFILE and OFILE)\nnumber of blocks that sg devices report from SCSI READ CAPACITY commands or that block\ndevices (or their partitions) report. Normal files are not probed for their size. If\nskip=SKIP or seek=SEEK are given and the count is derived (i.e. not explicitly given)\nthen the derived count is scaled back so that the copy will not overrun the device. If\nthe file name is a block device partition and COUNT is not given then the size of the\npartition rather than the size of the whole device is used. If COUNT is not given (or\ncount=-1) and cannot be derived then an error message is issued and no copy takes\nplace.\n\ndio={0|1}\ndefault is 0 which selects indirect (buffered) IO on sg devices. Value of 1 attempts\ndirect IO which, if not available, falls back to indirect IO and notes this at comple‐\ntion. If direct IO is selected and /proc/scsi/sg/allowdio has the value of 0 then a\nwarning is issued (and indirect IO is performed). For finer grain control use\n'iflag=dio' or 'oflag=dio'.\n\nibs=BS if given must be the same as BS given to 'bs=' option.\n\nif=IFILE\nread from IFILE instead of stdin. If IFILE is '-' then stdin is read. Starts reading\nat the beginning of IFILE unless SKIP is given.\n\niflag=FLAGS\nwhere FLAGS is a comma separated list of one or more flags outlined below. These\nflags are associated with IFILE and are ignored when IFILE is stdin.\n\nobs=BS if given must be the same as BS given to 'bs=' option.\n\nodir={0|1}\nwhen set to one opens block devices (e.g. /dev/sda) with the ODIRECT flag. User mem‐\nory buffers are aligned to the page size when set. The default is 0 (i.e. the ODIRECT\nflag is not used). Has no effect on sg, normal or raw files. If blksgio is also set\nthen both are honoured: block devices are opened with the ODIRECT flag and SCSI com‐\nmands are issued via the SGIO ioctl.\n\nof=OFILE\nwrite to OFILE instead of stdout. If OFILE is '-' then writes to stdout. If OFILE is\n/dev/null then no actual writes are performed. If OFILE is '.' (period) then it is\ntreated the same way as /dev/null (this is a shorthand notation). If OFILE exists then\nit is not truncated; it is overwritten from the start of OFILE unless 'oflag=append'\nor SEEK is given.\n\nof2=OFILE2\nwrite output to OFILE2. The default action is not to do this additional write (i.e.\nwhen this option is not given). OFILE2 is assumed to be a normal file or a fifo (i.e.\na named pipe). OFILE2 is opened for writing, created if necessary, and closed at the\nend of the transfer. If OFILE2 is a fifo (named pipe) then some other command should\nbe consuming that data (e.g. 'md5sum OFILE2'), otherwise this utility will block.\n\noflag=FLAGS\nwhere FLAGS is a comma separated list of one or more flags outlined below. These\nflags are associated with OFILE and are ignored when OFILE is /dev/null, '.' (period),\nor stdout.\n\nretries=RETR\nsometimes retries at the host are useful, for example when there is a transport error.\nWhen RETR is greater than zero then SCSI READs and WRITEs are retried on error, RETR\ntimes. Default value is zero.\n\nseek=SEEK\nstart writing SEEK bs-sized blocks from the start of OFILE. Default is block 0 (i.e.\nstart of file).\n\nskip=SKIP\nstart reading SKIP bs-sized blocks from the start of IFILE. Default is block 0 (i.e.\nstart of file).\n\nsync={0|1}\nwhen 1, does SYNCHRONIZE CACHE command on OFILE at the end of the transfer. Only ac‐\ntive when OFILE is a sg device file name or a block device and 'blksgio=1' is given.\n\ntime={0|1}[,TO]\nwhen 1, times transfer and does throughput calculation, outputting the results (to\nstderr) at completion. When 0 (default) doesn't perform timing.\nIf that value is followed by a comma, then TO is the command timeout in seconds for\nSCSI READ, WRITE or VERIFY commands issued by this utility. The default is 60 sec‐\nonds.\n\nverbose=VERB\nas VERB increases so does the amount of debug output sent to stderr. Default value is\nzero which yields the minimum amount of debug output. A value of 1 reports extra in‐\nformation that is not repetitive. A value 2 reports cdbs and responses for SCSI com‐\nmands that are not repetitive (i.e. other that READ and WRITE). Error processing is\nnot considered repetitive. Values of 3 and 4 yield output for all SCSI commands (and\nUnix read() and write() calls) so there can be a lot of output. This only occurs for\nscsi generic (sg) devices and block devices when the 'blksgio=1' option is set.\n", "subsections": [ { "name": "-d --dry-run", "content": "does all the command line parsing and preparation but bypasses the actual copy or\nread. That preparation may include opening IFILE or OFILE to determine their lengths.\nThis option may be useful for testing the syntax of complex command line invocations\nin advance of executing them.\n", "flag": "-d", "long": "--dry-run" }, { "name": "-h --help", "content": "outputs usage message and exits.\n", "flag": "-h", "long": "--help" }, { "name": "-p --progress", "content": "this option causes a progress report to be output every two minutes until the copy is\ncomplete. After the copy is complete a line with \"completed\" is printed to distinguish\nthe final report from the prior progress reports. When used twice the progress report\nis every minute, when used three times the progress report is every 30 seconds.\n", "flag": "-p", "long": "--progress" }, { "name": "-v --verbose", "content": "when used once, this is equivalent to verbose=1. When used twice (e.g. \"-vv\") this is\nequivalent to verbose=2, etc.\n", "flag": "-v", "long": "--verbose" }, { "name": "-x --verify", "content": "do a verify operation (like Unix command cmp(1)) rather than a copy. Cannot be used\nwith \"oflag=sparse\". of=OFILE must be given and OFILE must be an sg device or a block\ndevice with \"oflag=sgio\" also given. Uses the SCSI VERIFY command with the BYTCHK\nfield set to 1. The VERIFY command is used instead of WRITE when this option is given.\nThere is no VERIFY(6) command. Stops on the first miscompare unless oflag=coe is\ngiven.\n", "flag": "-x", "long": "--verify" }, { "name": "-V --version", "content": "outputs version number information and exits.\n", "flag": "-V", "long": "--version" } ] }, "CONVERSIONS": { "content": "One or more conversions can be given to the \"conv=\" option. If more than one is given, they\nshould be comma separated. sgdd does not perform the traditional dd conversions (e.g. ASCII\nto EBCDIC). Recently added conversions overlap somewhat with the flags so some conversions\nare now supported by sgdd.\n\nnocreat\nthis conversion has the same effect as \"oflag=nocreat\", namely: OFILE must exist, it\nwill not be created.\n\nnoerror\nthis conversion is very close to \"iflag=coe\" and is treated as such. See the \"coe\"\nflag. Note that an error on OFILE will stop the copy.\n\nnotrunc\nthis conversion is accepted for compatibility with dd and ignored since the default\naction of this utility is not to truncate OFILE.\n\nnull has no affect, just a placeholder.\n\nsparse FreeBSD supports \"conv=sparse\" so the same syntax is supported in sgdd. See \"sparse\"\nin the FLAGS sections for more information.\n\nsync is ignored by sgdd. With dd it means supply zero fill (rather than skip) and is typi‐\ncally used like this \"conv=noerror,sync\" to have the same functionality as sgdd's\n\"iflag=coe\".\n", "subsections": [] }, "FLAGS": { "content": "Here is a list of flags and their meanings:\n\n00 this flag is only active with iflag= and when given replaces if=IFILE. If both are\ngiven an error is generated. The input will be a stream of zeros, similar to using\n\"if=/dev/zero\" alone (but a little quicker).\n\nappend causes the OAPPEND flag to be added to the open of OFILE. For regular files this will\nlead to data appended to the end of any existing data. Cannot be used together with\nthe seek=SEEK option as they conflict. The default action of this utility is to over‐\nwrite any existing data from the beginning of the file or, if SEEK is given, starting\nat block SEEK. Note that attempting to 'append' to a device file (e.g. a disk) will\nusually be ignored or may cause an error to be reported.\n\ncoe continue on error. Only active for sg devices and block devices that have the 'sgio'\nflag set. 'iflag=coe oflag=coe' and 'coe=1' are equivalent. Use this flag twice (e.g.\n'iflag=coe,coe') to have the same action as the 'coe=2'. A medium, hardware or blank\ncheck error while reading will re-read blocks prior to the bad block, then try to re‐\ncover the bad block, supplying zeros if that fails, and finally re-read the blocks af‐\nter the bad block. A medium, hardware or blank check error while writing is noted and\nignored. A miscompare sense key during a VERIFY command (i.e. --verify given) is noted\nand ignored when 'oflag=coe'. The recovery of the bad block when reading uses the SCSI\nREAD LONG command if 'coe' given twice or more (also with the command line option\n'coe=2'). Further, the READ LONG will set its CORRCT bit if 'coe' given thrice. SCSI\ndisks may automatically try and remap faulty sectors (see the AWRE and ARRE in the\nread write error recovery mode page (the sdparm utility can access and possibly change\nthese attributes)). Errors occurring on other files types will stop sgdd. Error mes‐\nsages are sent to stderr. This flag is similar to 'conv=noerror,sync' in the dd(1)\nutility. See note about READ LONG below.\n\ndio request the sg device node associated with this flag does direct IO. If direct IO is\nnot available, falls back to indirect IO and notes this at completion. If direct IO\nis selected and /proc/scsi/sg/allowdio has the value of 0 then a warning is issued\n(and indirect IO is performed).\n\ndirect causes the ODIRECT flag to be added to the open of IFILE and/or OFILE. This flag re‐\nquires some memory alignment on IO. Hence user memory buffers are aligned to the page\nsize. Has no effect on sg, normal or raw files. If 'iflag=sgio' and/or 'oflag=sgio' is\nalso set then both are honoured: block devices are opened with the ODIRECT flag and\nSCSI commands are issued via the SGIO ioctl.\n\ndpo set the DPO bit (disable page out) in SCSI READ and WRITE commands. Not supported for\n6 byte cdb variants of READ and WRITE. Indicates that data is unlikely to be required\nto stay in device (e.g. disk) cache. May speed media copy and/or cause a media copy to\nhave less impact on other device users.\n\ndsync causes the OSYNC flag to be added to the open of IFILE and/or OFILE. The 'd' is\nprepended to lower confusion with the 'sync=0|1' option which has another action (i.e.\na synchronisation to media at the end of the transfer).\n\nexcl causes the OEXCL flag to be added to the open of IFILE and/or OFILE.\n\nff this flag is only active with iflag= and when given replaces if=IFILE. If both are\ngiven an error is generated. The input will be a stream of 0xff bytes (or all bits\nset).\n\nflock after opening the associated file (i.e. IFILE and/or OFILE) an attempt is made to get\nan advisory exclusive lock with the flock() system call. The flock arguments are\n\"FLOCKEX | FLOCKNB\" which will cause the lock to be taken if available else a \"tem‐\nporarily unavailable\" error is generated. An exit status of 90 is produced in the lat‐\nter case and no copy is done.\n\nfua causes the FUA (force unit access) bit to be set in SCSI READ and/or WRITE commands.\nThis only has an effect with sg devices or block devices that have the 'sgio' flag\nset. The 6 byte variants of the SCSI READ and WRITE commands do not support the FUA\nbit.\n\nnocache\nuse posixfadvise() to advise corresponding file there is no need to fill the file\nbuffer with recently read or written blocks.\n\nnocreat\nthis flag is only active in oflag=FLAGS. If present then OFILE will be opened if it\nexists. If OFILE doesn't exist then an error is generated. Without this flag a regular\n(empty) file named OFILE will be created (and then filled). For production quality\nscripts where OFILE is a device node (e.g. '/dev/sdc') this flag is recommended. It\nguards against the remote possibility of 'dev/sdc' disappearing temporarily (e.g. a\nUSB memory key removed) resulting in a large regular file called '/dev/sdc' being cre‐\nated.\n\nnull has no affect, just a placeholder.\n\nrandom this flag is only active with iflag= and when given replaces if=IFILE. If both are\ngiven an error is generated. The input will be a stream of pseudo random bytes. The\nLinux getrandom(2) system call is used to create a seed and thereadter mrand48r(3) is\nused to generate a pseudo random sequence, 4 bytes at a time. The quality of the ran‐\ndomness can be viewed with the ent(1) utility. This is not a high quality random num‐\nber generator, it is built for speed, not quality. One application is checking the\ncorrectness of the copy and verify operations of this utility.\n\nsgio causes block devices to be accessed via the SGIO ioctl rather than standard UNIX\nread() and write() commands. When the SGIO ioctl is used the SCSI READ and WRITE com‐\nmands are used directly to move data. sg devices always use the SGIO ioctl. This flag\noffers finer grain control compared to the otherwise identical 'blksgio=1' option.\n\nsparse after each BS * BPT byte segment is read from the input, it is checked for being all\nzeros. If so, nothing is written to the output file unless this is the last segment of\nthe transfer. This flag is only active with the oflag option. It cannot be used when\nthe output is not seekable (e.g. stdout). It is ignored if the output file is\n/dev/null . Note that this utility does not remove the OFILE prior to starting to\nwrite to it. Hence it may be advantageous to manually remove the OFILE if it is large\nprior to using oflag=sparse. The last segment is always written so regular files will\nshow the same length and so programs like md5sum and sha1sum will generate the same\nvalue regardless of whether oflag=sparse is given or not. This option may be used when\nthe OFILE is a raw device but is probably only useful if the device is known to con‐\ntain zeros (e.g. a SCSI disk after a FORMAT command).\n", "subsections": [] }, "RETIRED OPTIONS": { "content": "Here are some retired options that are still present:\n\nappend=0 | 1\nwhen set, equivalent to 'oflag=append'. When clear the action is to overwrite the ex‐\nisting file (if it exists); this is the default. See the 'append' flag.\n\nfua=0 | 1 | 2 | 3\nforce unit access bit. When 3, fua is set on both IFILE and OFILE; when 2, fua is set\non IFILE;, when 1, fua is set on OFILE; when 0 (default), fua is cleared on both. See\nthe 'fua' flag.\n", "subsections": [] }, "NOTES": { "content": "Block devices (e.g. /dev/sda and /dev/hda) can be given for IFILE. If neither '-iflag=di‐\nrect', 'iflag=sgio' nor 'blksgio=1' is given then normal block IO involving buffering and\ncaching is performed. If only '-iflag=direct' is given then the buffering and caching is by‐\npassed (this is applicable to both SCSI devices and ATA disks). If 'iflag=sgio' or\n'blksgio=1' is given then the SGIO ioctl is used on the given file causing SCSI commands to\nbe sent to the device and that also bypasses most of the actions performed by the block layer\n(this is only applicable to SCSI devices, not ATA disks). The same applies for block devices\ngiven for OFILE.\n\nVarious numeric arguments (e.g. SKIP) may include multiplicative suffixes or be given in\nhexadecimal. See the \"NUMERIC ARGUMENTS\" section in the sg3utils(8) man page.\n\nThe COUNT, SKIP and SEEK arguments can take 64 bit values (i.e. very big numbers). Other val‐\nues are limited to what can fit in a signed 32 bit number.\n\nData usually gets to the user space in a 2 stage process: first the SCSI adapter DMAs into\nkernel buffers and then the sg driver copies this data into user memory (write operations re‐\nverse this sequence). This is called \"indirect IO\" and there is a 'dio' option to select\n\"direct IO\" which will DMA directly into user memory. Due to some issues \"direct IO\" is dis‐\nabled in the sg driver and needs a configuration change to activate it. This is typically\ndone with 'echo 1 > /proc/scsi/sg/allowdio'.\n\nAll informative, warning and error output is sent to stderr so that dd's output file can be\nstdout and remain unpolluted. If no options are given, then the usage message is output and\nnothing else happens.\n\nEven if READ LONG succeeds on a \"bad\" block when 'coe=2' (or 'coe=3') is given, the recovered\ndata may not be useful. There are no guarantees that the user data will appear \"as is\" in the\nfirst 512 bytes.\n\nA raw device must be bound to a block device prior to using sgdd. See raw(8) for more in‐\nformation about binding raw devices. To be safe, the sg device mapping to SCSI block devices\nshould be checked with 'cat /proc/scsi/scsi', or sgmap before use.\n\nDisk partition information can often be found with fdisk(8) [the \"-ul\" argument is useful in\nthis respect].\n\nFor sg devices (and block devices when blksgio=1 is given) this utility issues SCSI READ and\nWRITE (SBC) commands which are appropriate for disks and reading from CD/DVD/HD-DVD/BD\ndrives. Those commands are not formatted correctly for tape devices so sgdd should not be\nused on tape devices. If the largest block address of the requested transfer exceeds a 32 bit\nblock number (i.e 0xffff) then a warning is issued and the sg device is accessed via SCSI\nREAD(16) and WRITE(16) commands.\n\nThe attributes of a block device (partition) are ignored when 'blksgio=1' is used. Hence the\nwhole device is read (rather than just the second partition) by this invocation:\n\nsgdd if=/dev/sdb2 blksgio=1 of=t bs=512\n", "subsections": [] }, "EXAMPLES": { "content": "Looks quite similar in usage to dd:\n\nsgdd if=/dev/sg0 of=t bs=512 count=1MB\n\nThis will copy 1 million 512 byte blocks from the device associated with /dev/sg0 (which\nshould have 512 byte blocks) to a file called t. Assuming /dev/sda and /dev/sg0 are the same\ndevice then the above is equivalent to:\n\ndd if=/dev/sda iflag=direct of=t bs=512 count=1000000\n\nalthough dd's speed may improve if bs was larger and count was suitably reduced. The use of\nthe 'iflag=direct' option bypasses the buffering and caching that is usually done on a block\ndevice.\n\nUsing a raw device to do something similar on a ATA disk:\n\nraw /dev/raw/raw1 /dev/hda\nsgdd if=/dev/raw/raw1 of=t bs=512 count=1MB\n\nTo copy a SCSI disk partition to an ATA disk partition:\n\nraw /dev/raw/raw2 /dev/hda3\nsgdd if=/dev/sg0 skip=10123456 of=/dev/raw/raw2 bs=512\n\nThis assumes a valid partition is found on the SCSI disk at the given skip block address\n(past the 5 GB point of that disk) and that the partition goes to the end of the SCSI disk.\nAn explicit count is probably a safer option. The partition is copied to /dev/hda3 which is\nan offset into the ATA disk /dev/hda . The exact number of blocks read from /dev/sg0 are\nwritten to /dev/hda (i.e. no padding).\n\nTo time a streaming read of the first 1 GB (2 30 bytes) on a disk this utility could be\nused:\n\nsgdd if=/dev/sg0 of=/dev/null bs=512 count=2m time=1\n\nOn completion this will output a line like: \"time to transfer data was 18.779506 secs, 57.18\nMB/sec\". The \"MB/sec\" in this case is 1,000,000 bytes per second.\n\nThe 'of2=' option can be used to copy data and take a md5sum of it without needing to re-read\nthe data:\n\nmkfifo fif\nmd5sum fif &\nsgdd if=/dev/sg3 iflag=coe of=sg3.img oflag=sparse of2=fif bs=512\n\nThis will image /dev/sg3 (e.g. an unmounted disk) and place the contents in the (sparse) file\nsg3.img . Without re-reading the data it will also perform a md5sum calculation on the image.\n", "subsections": [] }, "SIGNALS": { "content": "The signal handling has been borrowed from dd: SIGINT, SIGQUIT and SIGPIPE output the number\nof remaining blocks to be transferred and the records in + out counts; then they have their\ndefault action. SIGUSR1 causes the same information to be output yet the copy continues.\nAll output caused by signals is sent to stderr.\n", "subsections": [] }, "EXIT STATUS": { "content": "The exit status of sgdd is 0 when it is successful. Otherwise see the sg3utils(8) man page.\nSince this utility works at a higher level than individual commands, and there are 'coe' and\n'retries' flags, individual SCSI command failures do not necessary cause the process to exit.\n\nAn additional exit status of 90 is generated if the flock flag is given and some other\nprocess holds the advisory exclusive lock.\n", "subsections": [] }, "AUTHORS": { "content": "Written by Douglas Gilbert and Peter Allworth.\n", "subsections": [] }, "REPORTING BUGS": { "content": "Report bugs to .\n", "subsections": [] }, "COPYRIGHT": { "content": "Copyright © 2000-2021 Douglas Gilbert\nThis software is distributed under the GPL version 2. There is NO warranty; not even for MER‐\nCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.\n", "subsections": [] }, "SEE ALSO": { "content": "cmp(1)\n\nThere is a web page discussing sgdd at http://sg.danny.cz/sg/sgdd.html\n\nA POSIX threads version of this utility called sgpdd is in the sg3utils package. Another\nversion from that package is called sgmdd and it uses memory mapped IO to speed transfers\nfrom sg devices.\n\nThe lmbench package contains lmdd which is also interesting. For moving data to and from\ntapes see dt which is found at http://www.scsifaq.org/RMillerTools/index.html\n\nTo change mode parameters that effect a SCSI device's caching and error recovery see sd‐‐", "subsections": [ { "name": "parm(sdparm)", "content": "To verify the data on the media or to verify it against some other copy of the data see\nsgverify(sg3utils)\n\nSee also raw(8), dd(1), ddrescue(GNU), ddpt\n\n\n\nsg3utils-1.46 March 2021 SGDD(8)" } ] } }, "summary": "sgdd - copy data to and from files and devices, especially SCSI devices", "flags": [ { "flag": "-d", "long": "--dry-run", "arg": null, "description": "does all the command line parsing and preparation but bypasses the actual copy or read. That preparation may include opening IFILE or OFILE to determine their lengths. This option may be useful for testing the syntax of complex command line invocations in advance of executing them." }, { "flag": "-h", "long": "--help", "arg": null, "description": "outputs usage message and exits." }, { "flag": "-p", "long": "--progress", "arg": null, "description": "this option causes a progress report to be output every two minutes until the copy is complete. After the copy is complete a line with \"completed\" is printed to distinguish the final report from the prior progress reports. When used twice the progress report is every minute, when used three times the progress report is every 30 seconds." }, { "flag": "-v", "long": "--verbose", "arg": null, "description": "when used once, this is equivalent to verbose=1. When used twice (e.g. \"-vv\") this is equivalent to verbose=2, etc." }, { "flag": "-x", "long": "--verify", "arg": null, "description": "do a verify operation (like Unix command cmp(1)) rather than a copy. Cannot be used with \"oflag=sparse\". of=OFILE must be given and OFILE must be an sg device or a block device with \"oflag=sgio\" also given. Uses the SCSI VERIFY command with the BYTCHK field set to 1. The VERIFY command is used instead of WRITE when this option is given. There is no VERIFY(6) command. Stops on the first miscompare unless oflag=coe is given." }, { "flag": "-V", "long": "--version", "arg": null, "description": "outputs version number information and exits." } ], "examples": [ "Looks quite similar in usage to dd:", "sgdd if=/dev/sg0 of=t bs=512 count=1MB", "This will copy 1 million 512 byte blocks from the device associated with /dev/sg0 (which", "should have 512 byte blocks) to a file called t. Assuming /dev/sda and /dev/sg0 are the same", "device then the above is equivalent to:", "dd if=/dev/sda iflag=direct of=t bs=512 count=1000000", "although dd's speed may improve if bs was larger and count was suitably reduced. The use of", "the 'iflag=direct' option bypasses the buffering and caching that is usually done on a block", "device.", "Using a raw device to do something similar on a ATA disk:", "raw /dev/raw/raw1 /dev/hda", "sgdd if=/dev/raw/raw1 of=t bs=512 count=1MB", "To copy a SCSI disk partition to an ATA disk partition:", "raw /dev/raw/raw2 /dev/hda3", "sgdd if=/dev/sg0 skip=10123456 of=/dev/raw/raw2 bs=512", "This assumes a valid partition is found on the SCSI disk at the given skip block address", "(past the 5 GB point of that disk) and that the partition goes to the end of the SCSI disk.", "An explicit count is probably a safer option. The partition is copied to /dev/hda3 which is", "an offset into the ATA disk /dev/hda . The exact number of blocks read from /dev/sg0 are", "written to /dev/hda (i.e. no padding).", "To time a streaming read of the first 1 GB (2 30 bytes) on a disk this utility could be", "used:", "sgdd if=/dev/sg0 of=/dev/null bs=512 count=2m time=1", "On completion this will output a line like: \"time to transfer data was 18.779506 secs, 57.18", "MB/sec\". The \"MB/sec\" in this case is 1,000,000 bytes per second.", "The 'of2=' option can be used to copy data and take a md5sum of it without needing to re-read", "the data:", "mkfifo fif", "md5sum fif &", "sgdd if=/dev/sg3 iflag=coe of=sg3.img oflag=sparse of2=fif bs=512", "This will image /dev/sg3 (e.g. an unmounted disk) and place the contents in the (sparse) file", "sg3.img . Without re-reading the data it will also perform a md5sum calculation on the image." ], "see_also": [ { "name": "cmp", "section": "1", "url": "https://www.chedong.com/phpMan.php/man/cmp/1/json" }, { "name": "sgverify", "section": "sg3utils", "url": "https://www.chedong.com/phpMan.php/man/sgverify/sg3utils/json" }, { "name": "raw", "section": "8", "url": "https://www.chedong.com/phpMan.php/man/raw/8/json" }, { "name": "dd", "section": "1", "url": "https://www.chedong.com/phpMan.php/man/dd/1/json" }, { "name": "ddrescue", "section": "GNU", "url": "https://www.chedong.com/phpMan.php/man/ddrescue/GNU/json" } ] }