The outqueue_current PDU might also be in waitpdu if it does not have
ISCSI_PDU_DELETE_WHEN_SENT. outqueue_current is freed after the
waitpdu list (for reconnect or defer_reconnect), or sometimes not
considered at all (for cancel), and this can cause a dangling pointer.
Keep outqueue_current up to date when a PDU is freed. A bit hacky,
but it avoids touching code all over the place.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
if the rejected packet is a NOP-Out it is legal
that it has no callback. In this case we end
up in a segfault.
Signed-off-by: Peter Lieven <pl@kamp.de>
after a target has send the async target requests logout event it may
reject each request with a waiting for logout reason. Catch these
rejects if an outstanding request and the logout event overlap.
Signed-off-by: Peter Lieven <pl@kamp.de>
this adds progress output, 16 byte CDB support, additional checks and user
settable max_in_flight and blocks_per_io.
Signed-off-by: Peter Lieven <pl@kamp.de>
1.10 and 1.11 both introduced incompatible changes to the
libiscsi ABI. Please do not do that, and feel free to Cc
me on any patch that touches include/. In the meanwhile,
bump the soname of the library. I suggest releasing 1.11.1
with this change for distros.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
- writesame10_unmap_until_end, writesame16_unmap, writesame16_unmap_until_end were doing an CU_ASSERT *PER-BYTE* in the verification phase,
which was very CPU-intensive. This change uses memcmp on a whole block which finishes much quicker.
This patch uses appropriate format string specifiers for uint64_t
types taken from inttypes.h, instead of using %lu, because %lu is
wrong at least on 32bit architecture.
Also, cast size_t expressions to unsigned long, again, because on at least
32bit architectures, size_t is int, not long.
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
Cast ptrdiff (diff between two pointers) to long to fix printfs.
This fixes FTBFS on 32bit architectures.
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
Rename the macros for managing the linked lists from SLIST_* to ISCSI_LIST_*
to avoid a clash on *BSD which already have other macros SLIST_*
Signed-off-by: Ronnie Sahlberg <ronniesahlberg@gmail.com>
This version contains changes for:
- Fix memory leaks in persistent reserve out.
- Fix various leaks in the testsuite.
- Add additional READ16 tests
- Better logging when reconnect fails.
- Discovery and Login fixes.
- Various fixes found by coverity.
- WRITESAME test updates.
- Add XML oputput support for the testsuite.
and various other things.
Signed-off-by: Ronnie Sahlberg <ronniesahlberg@gmail.com>
The outqueue_current PDU might also be in waitpdu if it does not have
ISCSI_PDU_DELETE_WHEN_SENT. outqueue_current is freed after the
waitpdu list (for reconnect or defer_reconnect), or sometimes not
considered at all (for cancel), and this can cause a dangling pointer.
Keep outqueue_current up to date when a PDU is freed. A bit hacky,
but it avoids touching code all over the place.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Allow specifying a --test <file> in which case the tests listed in the file
will be executed.
This makes it easier than having to specify a potentially huge number
of tests on the command line.
Paolo Bonzini