Reduce the size of struct iscsi_context by reordering the members of this
data structure. Additionally, change the rdma_ack_timeout value from
'unsigned char' into 'uint8_t' to make it clear that this variable
represents an integer.
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Since 2c1619edef61a03cb516efaa81750784c3071d10 for linux kernel and
55843c4ab8f559679d28c559cc4d681836be769b for rdma-core, rdma cma
supports RDMA_OPTION_ID_ACK_TIMEOUT. It's useful for RDMA out of
sequence case. Because this feature is added recently, we have to
check this in autogen.sh before building source code.
Depend on production enviroument, tunning rdma ack timeout could get
the best performance. Suggested by Bart and Ronnie, instead of using
a fixed timeout value, add two methods to set rdma ack timeout value.
1, add URL variable 'LIBISCSI_RDMA_ACK_TIMEOUT'. This could works
for a specified connection.
2, add env argument 'LIBISCSI_RDMA_ACK_TIMEOUT'. This works as a
common setting for all the connection of a process.
Test under different packet loss rate and different ack timeout, run
fio (iodepth=1) in a guest os, I got this result:
latency under packet loss rate 0.00001:
timeout 19: avg 170.22, pct99.9 215
timeout 10: avg 160.08, pct99.9 215
timeout 8 : avg 146.39, pct99.9 177
timeout 7 : avg 148.37, pct99.9 211
latency under packet loss rate 0.0001:
timeout 19: avg 949.23, pct99.9 306
timeout 10: avg 818.53, pct99.9 378
timeout 8 : avg 615.84, pct99.9 189
timeout 7 : avg 618.89, pct99.9 310
Base on this test report, setting ack timeout to 8(1048.576 usec) is
a good choice in my test enviroument.
Signed-off-by: zhenwei pi <pizhenwei@bytedance.com>
Declare dynamically allocated strings as 'char *' instead of 'const char *'.
Remove the discard_const() macro. Do not test whether or not a pointer is
NULL before calling free() because it is allowed to pass NULL to free().
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
The buffer is memcopied into the PDU. const makes it a little clearer
that the caller isn't handing over ownership.
Signed-off-by: David Disseldorp <ddiss@suse.de>
The ISCSI_HEADER_SIZE macro accesses iscsi->header_size, so pass it in
as a parameter to make it easier to follow callers.
Signed-off-by: David Disseldorp <ddiss@suse.de>
In our use at Datto, we have come across several issues related to the
automatic reconnect logic (mainly its interaction with POLLHUP). This
allows us to disable the functionality, at the expense of writing our
own reconnect logic.
Related: #241
Introduce a helper exported from lib/pdu.c which cancels all pdus for a
given context. This patch eliminates repeated code from various other
files which have the same purpose. The only functional difference is
that the cancellation done from iscsi-command.c was (incorrectly) not
checking for iscsi->is_loggedin before issuing callbacks.
Signed-off-by: Felipe Franciosi <felipe@nutanix.com>
Using WIN32 depends on the build environment defining the variable.
_WIN32 is a predefined MSVC macro and is always available.
Signed-off-by: Tim Crawford <crawfxrd@gmail.com>
Win32 has been rotting for a while. This patch adds vs17 build files
as well as fixing up all build errors that have accumulated.
There are still build warnings but those can be addressed in a followup
patch.
Signed-off-by: Ronnie Sahlberg <ronniesahlberg@gmail.com>
iscsi_read_from_socket can currently only read one PDU in each iscsi_service invocation even
if there is more data available on the socket. This patch reads all PDUs until the socket
would block. It enqueues all complete read PDUs and then processes them in order of arrival.
Signed-off-by: Peter Lieven <pl@kamp.de>
This splits a transport into static driver specific functions for the common
iscsi commands. Optionally, a driver specific opaque memory is introduced
which is currently only used by iSER transport.
Last a lot of functions changed to static.
Signed-off-by: Peter Lieven <pl@kamp.de>
This commit includes all iSER implementation in libscsi
library and utilities.
Also, adding iser option in url.
Change-Id: I55ca8a9d4db802e72eb991061260dbb0bd0ef9ba
Signed-off-by: Roy Shterman <roysh@mellanox.com>
future iSER implementation will include different implementations
for all socket relative function. in iSER we get event only when
there is new entry in completion queue opposed to TCP that we get event
when we can write to the socket.
1. iscsi_get_fd -
TCP - returns socket fd.
ISER - returns completion queue channel fd.
2. iscsi_service -
TCP - processing the event type got from the socket
and handles it.
ISER - rearming the event mechanism in the completion queue
and polling all available completion queue entries for
process.
3. iscsi_which_events -
TCP - returns which type of event the library is waiting for
(Read, Write or both).
ISER - in iSER we are waiting only for POLLIN event, hence this
function always returns POLLIN.
Signed-off-by: Roy Shterman <roysh@mellanox.com>
all library: change disconnect to iscsi->t->disconnect
1. In TCP we need only to put -1 in fd and we don't
have more transport resources. In future iSER we will need to
clean resources and destroy the rdma connection.
Signed-off-by: Roy Shterman <roysh@mellanox.com>
socket: need to malloc hdr
include/iscsi-private: changing iscsi_in_pdu hdr to char*
instead of static array for more convinient iser
pdu creation.
To use iscsi_in_pdu in iSER without making a copy of it
we need to change hdr to pointer from static array,
Because of that, iscsi_tcp flow need to do szmalloc (small zero malloc)
hdr when creating new iscsi_in_pdu. iscsi_in_pdu is being malloced
once per each received pdu. This change is reducing iscsi_in_pdu struct
size but adding extra allocation of the same size we reduced
from the struct.
Signed-off-by: Roy Shterman <roysh@mellanox.com>
include/iscsi-private: adding queue_pdu in transport function pointers
struct
include/iscsi: declaration of tcp_queue_pdu function
socket: adding queue_pdu function to transport initialization
all_library: changing iscsi_queue_pdu into iscsi->t->queue_pdu
Signed-off-by: Roy Shterman <roysh@mellanox.com>
socket: adding tcp_connect function and implement it
in the last common part of iSER and TCP in iscsi_connect_async
Signed-off-by: Roy Shterman <roysh@mellanox.com>
lib/init: initializing connection transport
lib/socket: Adding function to initialize tcp transport.
future commits will include adding iSER transport layer,
so each transport option will has it's own template of functions.
all_utils: All utils stay the same, in future iSER commits in case of
iSER transport we will override tcp_transport in
iscsi_context.
connect: adding transport initialization when creating context
include/iscsi-private: Adding iscsi_transport attribute in iscsi_context
Signed-off-by: Roy Shterman <roysh@mellanox.com>
Make iscsi_init_tcp_transport private
Signed-off-by: Ronnie Sahlberg <ronniesahlberg@gmail.com>
we use the flag ISCSI_PDU_NO_CALLBACK and pdu->callback simultaneously, but
check only for one of them in various places. So drop ISCSI_PDU_NO_CALLBACK
and check for pdu->callback != NULL instead.
All PDUs that carried this flag have pdu->callback set to NULL.
Signed-off-by: Peter Lieven <pl@kamp.de>
This seriously breaks qemu NOP timeouts and probably other things.
The reason is that the
define ISCSI_PDU_ERROR_ON_RECONNECT 0x00000016
is masking bits 0x2 and 0x4 as well.
Correctly it should read:
define ISCSI_PDU_ERROR_ON_RECONNECT 0x00000010
However, the better solution for this approach is invoke all callbacks
of PDUs which carry the ISCSI_PDU_DROP_ON_RECONNECT flag. This will
make sure that callbacks of whatever sync tasks are invoked.
This reverts commit 0407cf6aed.
Certain iSCSI commands such as NOP and LOGOUT commands are discarded instead
of re-queued when we have a session failure and reconnect.
Change the LOGOUT command to instead fail with SCSI_STATUS_ERROR when this
happens.
Otherwise, IF we are in iscsi_logout_sync() and we get a session failure
at the same-ish time we may end up automatically re-connecting the
session, but since we have discarded the logout command we will never
get a reply and will hang indefinitely in the event loop for synchronous
commands.
Arguably, we could also just return SCSI_STATUS_GOOD here since
when we perform a logout, we probably don't care too much about how we
disconnected from the server, only that we did disconnect from the server.
That is academic anyway since this only affects the sync API which is only
meant for trivial applications, which will likely not inspect the result
and just do a:
...
iscsi_logout_sync()
iscsi_destroy_context()
...
anyway.
Signed-off-by: Ronnie Sahlberg <ronniesahlberg@gmail.com>
Normal applications want the current behaviour where we have the library
consume any/all of the UnitAttentions that the target may have queued on the
initial connection, but when we reconnect the session after a failure the
library will pass all the UAs back to the application to process.
Some applications, such as the test suite or really trivial applications
might not want to have to deal with handling of UAs and just "make it work".
Those applications can now request that upon any reconnection of the session
that libiscsi will automatically consume any and all UAs and hide them from
the application.
Signed-off-by: Ronnie Sahlberg <ronniesahlberg@gmail.com>
this variable was introduced for iscsi-test-cu only. This patch
makes it a generic environment variable that can be set per context.
Signed-off-by: Peter Lieven <pl@kamp.de>
If we are reconnecting and the connection is already established we
end up looping forever if there is a socket error before we finish
the login process. Fix this by invoking the reconnect callback if
there is an error and no new reconnect is scheduled.
Signed-off-by: Peter Lieven <pl@kamp.de>
This is a bug that has been there a long time.
When we reconnect and requeue a PDU we must reset the iovectors
for the task. Otherwise, any partially sent/received data when the
command is reconnected would end up containing garbage.
Signed-off-by: Ronnie Sahlberg <ronniesahlberg@gmail.com>
Rework the reconnect logic so we just call iscsi_scsi_command_async()
for the scsi commands we are re-quining instead of poking into the
private fields of the structures themself.
Signed-off-by: Ronnie Sahlberg <ronniesahlberg@gmail.com>
Remove iscsi_allocate_pdu() which is just a wrapper.
Rename iscsi_allocate_pdu_with_itt_flags() to iscsi_allocate_pdu()
and update all callers.
This only removes a wrapper function and contains no logic changes.
Signed-off-by: Ronnie Sahlberg <ronniesahlberg@gmail.com>
We write unsolicited data-out PDUs from two places;
when we originally write the command in iscsi_scsi_command_async()
but also when we re-queue the PDUs during a session reconnect.
The re-queuing during the session re-connect was recently (almost) fixed
but was still buggy in that it did not correctly clamp the amount of written
data as per first burst length restriction.
This attempts to fix that.
Signed-off-by: Ronnie Sahlberg <ronniesahlberg@gmail.com>
If we have writes that do not have the Final bit set during reconnect
we must send out any missing data-out PDU.
Signed-off-by: Ronnie Sahlberg <ronniesahlberg@gmail.com>
