Ronnie Sahlberg
ee689ee3fb
TESTS: Update the protection tests
2013-05-12 08:06:08 -07:00
Ronnie Sahlberg
67676d4756
TESTS: *protect field set to non-zero are only guaranteed to fail if the device does not support protection information
2013-05-11 17:27:35 -07:00
Ronnie Sahlberg
5135836ee5
TESTS: keep the unmarshalled std inq structure around
2013-05-11 17:07:59 -07:00
Ronnie Sahlberg
772d220217
Add a version descriptor
2013-05-11 16:03:23 -07:00
Ronnie Sahlberg
3a31759fa4
Merge pull request #62 from bonzini/master
...
MD5 updates
2013-05-03 06:36:35 -07:00
Paolo Bonzini
bcb2950d8d
use libgcrypt for MD5
...
This makes sure that CHAP authentication is disabled if the system
is running in FIPS 140-2 mode. MD5 is not a secure algorithm according
to the standard.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com >
2013-05-03 12:51:15 +02:00
Paolo Bonzini
cc742279d0
fix bug in md5 code
2013-05-03 12:50:58 +02:00
Ronnie Sahlberg
d38cf7cfcd
TESTS: since we track expcmdsn we dont need to keep the old value around in a static variable
2013-04-30 19:17:59 -07:00
Ronnie Sahlberg
b7dd6b533b
TEST: Add a test that is cmdsn is too low the target just ignores the pdu
2013-04-30 19:05:23 -07:00
Ronnie Sahlberg
325360789e
TEST: Improve the test for cmdsn too high with more documentation
...
and verify that the target did ignore the command and the session is still good.
2013-04-30 18:46:26 -07:00
Ronnie Sahlberg
1a275bbecc
TEST: rename the cmdsn too high test
2013-04-30 18:32:09 -07:00
Ronnie Sahlberg
1165e06d19
Fix a bunch of typos
2013-04-29 21:02:40 -07:00
Ronnie Sahlberg
4a8d967541
Add support for synchronous command timeout.
...
Default to 0 meaning no timeout.
Implement a test for iSCS to test what happens if we send a command
with CMDSN being higher than the target allows.
In this case we dont strictly know what will happen, just that what should
NOT happen is the target responding with success.
But we have to be prepared for any kind of failure, including a timeout,
scsi sense, or even iscsi reject or session failure.
2013-04-29 20:42:33 -07:00
Ronnie Sahlberg
5bce69f1f3
TEST: rename the usb block device tests to usb-sbc
2013-04-27 20:39:30 -07:00
Ronnie Sahlberg
f8e317c355
TESTS: READ6 is optional(==obsolete)
2013-04-27 16:22:58 -07:00
Ronnie Sahlberg
94d048afae
TESTS: Persistent reserve in/out are optional
2013-04-27 14:09:19 -07:00
Ronnie Sahlberg
4222d5fcd7
TESTS: Add another level for the test hiearchy : Familt/Suite/Test
...
and add 4 basic families : ALL, SCSI, iSCSI and USB-MEMORY-STICK
2013-04-27 13:57:57 -07:00
Ronnie Sahlberg
9d44644c63
TESTS: Move the tests for iscsi residuals to its own suite
2013-04-27 10:15:30 -07:00
Ronnie Sahlberg
2627a8845e
TESTS: add a flag --usb so we can clamp all i/o to at most 120k
...
Add checks for maximum tranfer length to all commands that are limited
by BlockLimits/MTL
2013-04-27 09:59:59 -07:00
Ronnie Sahlberg
c866787dcb
TESTS: read12 is optional, should make test skip, not fail
2013-04-27 07:43:24 -07:00
Ronnie Sahlberg
227489d32b
TESTS: fix PREFETCH10/16 tests
2013-04-26 18:01:04 -07:00
Ronnie Sahlberg
1331656f85
TESTS: ORWRITE is optional, lack os support should skip the test not fail it
2013-04-26 17:52:55 -07:00
Ronnie Sahlberg
78e02737c6
TESTS: fix some bugs in the tests for INQUIRY
2013-04-26 17:43:06 -07:00
Ronnie Sahlberg
e33194d94a
TESTS: WRITEVERIFY10/12/16 are all optional.
...
Update the helpers to return -2/not supported on invalid operaction code.
Update all tests to accept that the opcode might be missing and abort the test without failing.
2013-04-25 17:41:13 -07:00
Ronnie Sahlberg
b5804ee499
TESTS: Update WRITE10/12 to allow it to fail with invalid operation code. Both these opcodes are Optional in all SBC standards.
2013-04-23 21:29:25 -07:00
Ronnie Sahlberg
ee666ae451
TESTS: WRITE16 is optional so update the tests to allow WRITE16 to be missing and just make the test be [SKIPPED]
2013-04-23 21:15:28 -07:00
Ronnie Sahlberg
d280ce8ad7
Create safe 16/32/64 bit accessors for reading from the datain buffer
...
and use it throughout the scsi lowlevel file.
We probably want a safe accessor for byte access to at some stage.
2013-04-22 22:11:53 -07:00
Ronnie Sahlberg
0a026e2da7
AROS: We need emulation of readv/writev on aros
2013-04-22 18:50:05 -07:00
Ronnie Sahlberg
322be0cf24
add missing version descriptor
2013-04-22 18:34:02 -07:00
Ronnie Sahlberg
555906e76e
Revert "Add SSC to the version descriptors."
...
This reverts commit 45f5fc2105 .
Conflicts:
include/scsi-lowlevel.h
2013-04-22 18:32:09 -07:00
Ronnie Sahlberg
e640fa3cde
missing , broke compile
2013-04-22 18:28:20 -07:00
Ronnie Sahlberg
45f5fc2105
Add SSC to the version descriptors.
...
I actually have a usb-stick that claims it supports SSC. I stongly doubt it.
2013-04-22 17:44:13 -07:00
Ronnie Sahlberg
284514b14e
fix typo
2013-04-22 17:38:10 -07:00
Ronnie Sahlberg
79f54c5e06
TESTS update all read16 tests to fail/skip on illegal opcode depending on whether the device claims support for it or not (mandatory in sbc-3, optional prior)
2013-04-22 17:31:50 -07:00
Ronnie Sahlberg
77ff1b1808
TESTS: make the "is readcapacity16 supported on this device" tests cleaner
2013-04-22 17:11:38 -07:00
Ronnie Sahlberg
af6d1c9c13
Add a test for mandatory opcodes for SBC devices.
...
Note that Read16/Readcapacity16 are mandatory on SBC3 but not prior.
Not all mandatory opcodes are tested for yet.
2013-04-21 14:50:41 -07:00
Ronnie Sahlberg
91a98d6b92
Add handling of inquiry version descriptors
2013-04-21 14:04:30 -07:00
Ronnie Sahlberg
b3ca40efbe
Get rid of the popt dependency
2013-04-21 10:25:30 -07:00
Ronnie Sahlberg
a048079c90
covnert the old test suite grom popt to getopt_long
2013-04-21 10:19:40 -07:00
Ronnie Sahlberg
ecfe404316
iscsi-support.c does not need popt.h
2013-04-21 10:03:34 -07:00
Ronnie Sahlberg
1c5d132ed0
convert iscsi-dd to use getopt_long()
2013-04-21 10:02:05 -07:00
Ronnie Sahlberg
f5d25c0157
ISCSI-READCAPACITY16 convert to getopt_long()
2013-04-21 09:58:15 -07:00
Ronnie Sahlberg
c3886cf9c4
ISCSI-INQ: convert to getopt_long instead of popt
2013-04-21 09:53:12 -07:00
Ronnie Sahlberg
c0d1c122c3
ISCSI-LS : Change from popt to the slightly less portable getopt_long
...
win32 and amiga/aros does not have getopt_long() so we will need a
replacement for the platforms where getopt_long() is missing.
2013-04-21 09:44:23 -07:00
Ronnie Sahlberg
622b38d679
Add support for AROS
2013-04-18 19:45:17 -07:00
Ronnie Sahlberg
402653b9f3
portability updates
...
add check if ipv6/sockaddr_in6 is availavble or not
add check for poll.h and only include when available
add includes for AROS
2013-04-18 19:43:36 -07:00
Ronnie Sahlberg
fbf58ead9f
TESTS: When task management functions fail, print the reason why it failed.
2013-03-27 17:25:47 -07:00
Ronnie Sahlberg
aac7ed29b8
SYNC Task mgmt functions. Return !0 status if the command completed successfully
...
but the target responded with non-zero response code.
2013-03-26 17:38:47 -07:00
Ronnie Sahlberg
d6f4c4d259
TESTS update reserve6 tests
2013-03-25 17:49:11 -07:00
Ronnie Sahlberg
2873b18bb7
TESTS: Add tests to verify that RELEASE6 is dropped on target/lun reset
2013-03-25 17:42:24 -07:00