summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* Preparing for version 2.1.12.1.1new-versionLee Duncan2020-02-263-2/+35
|
* Merge pull request #187 from ZhiqiangLiu26/masterLee Duncan2020-02-211-0/+4
|\ | | | | Check whether socket is opened successfully in find_vlan_dev func
| * Check whether socket is opened successfully in find_vlan_dev funcl004648062020-02-031-0/+4
| | | | | | | | | | | | In find_vlan_dev func, socket should be checked before used. Signed-off-by: Zhiqiang Liu <liuzhiqiang26@huawei.com>
* | Merge pull request #189 from gonzoleeman/fix-libopeniscsiusr-valgrindLee Duncan2020-02-201-4/+6
|\ \ | | | | | | Fix memory leaks in libopeniscsiusr/idbm.c
| * | Fix memory leaks in libopeniscsiusr/idbm.cLee Duncan2020-02-201-4/+6
| |/ | | | | | | | | Running "make check" in the libopeniscsiusr directory found two memory leaks. The fix was to free the memory.
* | Merge pull request #188 from gonzoleeman/update-regression-testLee Duncan2020-02-2011-10/+815
|\ \ | |/ |/| Update regression test
| * Allow sub-tests to be specified.Lee Duncan2020-02-202-21/+121
| | | | | | | | | | | | | | Using the new "-s" option, one can specify a single subtest or a range of subtests. Together with the blocksize arguments and being able to specify the actual test case if you wish, you can now run just a single test.
| * more things to doLee Duncan2020-02-201-4/+30
| |
| * Test code rearranged to make discovery work.Lee Duncan2020-02-206-122/+156
| | | | | | | | | | | | | | | | | | | | | | | | Now you can run: > python3 -m unittest discover to run all the tests, using discovery, or > ./test-open-iscsi -l to list the tests.
| * Add in "-V" for version infoLee Duncan2020-02-202-5/+16
| |
| * Update TODO listLee Duncan2020-02-202-0/+11
| |
| * 64 tests now working and passingLee Duncan2020-02-203-25/+68
| | | | | | | | | | Tested on Tumbleweed, and all tests pass so far. Run time is around 3 hours so far, for the whole suite.
| * First 32 tests working?Lee Duncan2020-02-194-70/+148
| | | | | | | | At least they are coded.
| * Beginning to get python tests set up.Lee Duncan2020-02-198-10/+512
|/ | | | | Created a framework. Replacing disktest with fio, sfdisk with parted, and moving to a PyUnit test framework.
* Merge pull request #186 from knakamu/testLee Duncan2020-01-311-0/+3
|\ | | | | Avoid logout of iscsi boot session
| * Avoid logout of iscsi boot sessionKiyotaka Nakamura2020-01-311-0/+3
|/
* Merge pull request #185 from gonzoleeman/fix-compiler-issuesLee Duncan2020-01-3031-132/+214
|\ | | | | Fix compiler issues
| * Fix 586 compiler issues.Lee Duncan2020-01-299-19/+27
| | | | | | | | | | | | | | Now that "-Werror" is turned on, 32-bit x86 compile problems are fatal, so fix them. *Should* be no functional change.
| * Updates to support gcc -fno-common option.Lee Duncan2020-01-286-6/+11
| | | | | | | | | | | | This meant cleaning up the definition of some global variables, so that they were only defined in one place and refered to as external elsewhere.
| * Enabled compiler checking options, fixed issues.Lee Duncan2020-01-2823-108/+177
|/ | | | | | Enabled -Wextra and -Werror, then fixed all the issues found. *Should* be no functional change.
* Revert "Out-of-bounds read: Overrunning array of 8 2-byte elements"Chris Leech2020-01-201-1/+1
| | | | | | It's been reported that this breaks IPv6 discovery with qedi controllers This reverts commit af48f0d68db48cd492d0e27e046fc011be236fd3.
* Merge pull request #183 from phmccarty/pc-fileChris Leech2020-01-081-1/+1
|\ | | | | Fix bug with libopeniscsiusr.pc
| * Fix bug with libopeniscsiusr.pcPatrick McCarty2020-01-081-1/+1
| | | | | | | | | | The library name is libopeniscsiusr, so the -l linker option should be `-lopeniscsiusr` instead of `-liscsiusr`.
* | Merge pull request #182 from cleech/chap_configChris Leech2020-01-0814-32/+294
|\ \ | |/ |/| configuration support for CHAP algorithms
| * configuration support for CHAP algorithmsChris Leech2019-12-1914-32/+294
|/ | | | | | | | Introduces support for preference lists in configuration files, and uses that for the 'node.session.auth.chap_algs' setting. This is also re-used for discovery authentication, rather than have two different configurations.
* Fixed Changelog message, first lineLee Duncan2019-11-141-1/+1
|
* Preparing for version 2.1.02.1.0Lee Duncan2019-11-143-2/+41
| | | | | | | | | | | Updated Changelog and two include files. Note that this bumps the minor version number from 0 to 1, and starts the "patch" version number over again at 0. See https://semver.org for information on semantic versioning.
* Fix version strings in ChangeLogLee Duncan2019-11-121-6/+6
| | | | | Some places list version number with "2.0-NUMBER", and some as "2.0.NUMBER". The latter is correct.
* Merge pull request #178 from open-iscsi/coverity_scanChris Leech2019-11-0417-54/+65
|\ | | | | first batch of changes to address Coverity scan issues
| * Uninitialized scalar value rcChris Leech2019-10-311-1/+1
| |
| * Resource leak: Variable raw going out of scope leaks the storage it points to.Chris Leech2019-10-311-0/+1
| |
| * Out-of-bounds read: Overrunning array of 8 2-byte elementsChris Leech2019-10-311-1/+1
| | | | | | | | | | This loop is executing 16 times (sizeof(struct ipv6_addr)) but then consumes 16-bits at a time (and moves ptr ahead two bytes).
| * Resource leak: Handle variable fd going out of scope leaks the handle.Chris Leech2019-10-311-0/+1
| |
| * fwparam_pcc mulitple resource leaksChris Leech2019-10-311-2/+1
| | | | | | | | | | Resource Leak: Variable mac_file going out of scope leaks the storage it points to. Resource Leak: Handle variable mac_fd going out of scope leaks the storage it points to.
| * iscsistart -b probably never worked with PPC OF parsing?Chris Leech2019-10-311-1/+0
| | | | | | | | | | | | Picked up by Coverity as a resource leak on the allocated context, callers are not expected allocation here (and the function would need to take a **boot_context if they were)
| * iscsi-iname remove unneeded temp bufferChris Leech2019-10-311-6/+1
| |
| * iscsi-iname: verify prefix length is at most 210 charactersChris Leech2019-10-311-0/+9
| | | | | | | | Don't know who might be trying to make really long IQNs ¯\_(ツ)_/¯
| * iscsi-iname: change default IQN prefixChris Leech2019-10-312-2/+2
| | | | | | | | | | We don't actually control the open-iscsi.org domain anymore, so change the prefix to reflect the use of open-iscsi.com
| * Out-of-bounds read: Overrunning array of 4 bytes at byte offset 7 by ↵Chris Leech2019-10-311-2/+2
| | | | | | | | dereferencing pointer
| * Resource leak: Handle variable fd going out of scope leaks the handle.Chris Leech2019-10-311-0/+1
| |
| * Resource leak: Handle variable fd going out of scope leaks the handle.Chris Leech2019-10-311-0/+2
| | | | | | | | | | | | This is comming up in places where the error handling assumes that on a failure there is no file descriptor to worry about. I don't think we need to keep fd around when indicating a connection error.
| * Resource leak: Variable matched_ses going out of scope leaks the storage it ↵Chris Leech2019-10-311-2/+2
| | | | | | | | points to.
| * Resource leak: Variable chap_info going out of scope leaks the storage it ↵Chris Leech2019-10-311-1/+1
| | | | | | | | points to.
| * Resource leak: Handle variable sockfd going out of scope leaks the handle.Chris Leech2019-10-311-1/+3
| |
| * Uninitialized scalar variable: Using uninitialized value number when calling ↵Chris Leech2019-10-311-1/+1
| | | | | | | | | | | | | | | | | | acl_text_to_number This was triggering from acl_text_to_number callsites, but the issue to me looks like the unneeded reading of the parameter num. The variable number is always going to be overwritten after this with the output of strtoul.
| * Uninitialized pointer read: Using uninitialized value ifaces.nextChris Leech2019-10-311-20/+11
| | | | | | | | | | | | | | | | On iscsi_context_new failure in main, this was jumping to out without the ifaces list head being initialized. Run through this file making use of LIST_HEAD for static initialization where possible.
| * Uninitialized scalar variableChris Leech2019-10-311-1/+1
| |
| * Buffer not null terminated: Calling strncpy with a maximum size argument on ↵Chris Leech2019-10-311-3/+3
| | | | | | | | destination array
| * Resource leak: Variable startup_cmd going out of scope leaks the storage it ↵Chris Leech2019-10-311-0/+2
| | | | | | | | point to.
| * Out-of-bounds access: Overrunning array value_listChris Leech2019-10-311-1/+3
| |
View Lorry Controller Status