1. Add a test script and expect script for testing forwarding.
2. Update portfwd to have a ready file option.
3. Fix echoserver error string, needed NL.
1. Fix a couple unused variable warnings.
2. In wolfSSH_AGENT_DefaultActions(), fix comparison to the result of
snprintf() treating normal result as an error. Reset the return code
for the error state of the socket() command. Better cleanup of agent
startup failures.
When running the build tests, the terminal size gets set to 0x0. When
then tests finish, things like vim start to misbehave until you change
your term window size.
1. Echoserver will not change the terminal size when in echo mode.
2. Do not update the modes when the echo server is not in echo mode.
3. Do not set the terminal update callback context when the echo server
is not in echo mode.
4. Update the guards around the above three items to match the guards for
the modes update function.
Historically, some systems have used <sys/errno.h> in place of <errno.h>, but the C standard,
all the way back to C89/Ansi C, specifies that the correct name is `errno.h` -
and most modern compilers create a sys/errno.h compatibility shim sys/errno.h -
but musl-gcc does not have this compatibility shim,
reasoning that sys/errno.h is non-standard, and fails to compile:
```
$ CC=musl-gcc ./configure --enable-static --disable-shared --enable-scp --enable-sftp --enable-shell --with-wolfssl=wolfssl/install-musl
(...)
$ make
make -j17 all-am
make[1]: Entering directory '/home/hans/projects/wolfssh'
CC examples/client/common.o
CC examples/client/client.o
CC src/libwolfssh_la-ssh.lo
CC src/libwolfssh_la-internal.lo
CC src/libwolfssh_la-log.lo
CC src/libwolfssh_la-port.lo
CC src/libwolfssh_la-io.lo
CC src/libwolfssh_la-wolfscp.lo
CC src/libwolfssh_la-wolfsftp.lo
CC examples/echoserver/echoserver.o
CC src/libwolfssh_la-wolfterm.lo
CC examples/sftpclient/sftpclient.o
CC examples/scpclient/scpclient.o
In file included from examples/echoserver/echoserver.c:91:
/usr/include/x86_64-linux-musl/sys/errno.h:1:2: error: #warning redirecting incorrect #include <sys/errno.h> to <errno.h> [-Werror=cpp]
1 | #warning redirecting incorrect #include <sys/errno.h> to <errno.h>
| ^~~~~~~
cc1: all warnings being treated as errors
make[1]: *** [Makefile:1452: examples/echoserver/echoserver.o] Error 1
make[1]: *** Waiting for unfinished jobs....
make[1]: Leaving directory '/home/hans/projects/wolfssh'
make: *** [Makefile:1020: all] Error 2
```
so, unless we need to support pre-C89 (35+ year old C) compilers,
we can just replace sys/errno.h with errno.h.
1. Modify all the test executable `CPPFLAGS` to only have
`-DNO_MAIN_LOOP` and `AM_CPPFLAGS`. This removes a lot of duplicate
setup of flags for each of the tests.
2. Delete the echoserver's `CFLAGS`. It is just setting the `CFLAGS` for
its compile to `AM_CFLAGS` and `AM_CPPFLAGS` which are included by
default.
3. Add the macro WS_MAYBE_UNUSED that may be the unused attribute.
4. If INLINE isn't set to some form of inline, set it to
WS_MAYBE_UNUSED.
5. Remove redundant ThreadStartNoJoin() macros from the testing header.
This adds the `diffie-hellman-group16-sha512` key exchange and
`hmac-sha2-512` mac support.
Echoserver can now take `-x` for key exchange and `-m` for mac setting,
and `-c` for cipher so that this can be used in the test suite.
This implements Keyboard-Interactive authentication.
Adds an additional callback set by `wolfSSH_KeyboarAuthPrompts()` which
will set a callback in the server to ask the application to provide the
prompt details for the client.
1. Fix echoserver's load_file function. Could potentially dereference
null if a file size is passed in, but using a null buf to get the
file's size only.
1. Add the code from wolfsshd that initially sets up the terminal
settings to the echoserver.
2. Add the WOLFSSH_TERM flag to the build of the echoserver.
1. In the echoserver, change the names of the defines for the
static memory values.
2. Modify the client common functions for keys and certs to take a heap
value for their allocations.
3. Update the client's use of the common key and cert function calls to
pass NULL for the heap.
4. Add a static memory description for the sftp client code.
1. Modify the echoserver to use a static memory pool when using the flag
WOLFSSH_STATIC_MEMORY. Uses NULL otherwise. The WOLFSSH_CTX is
created with the pool and subsequent allocations for the WOLFSSH
server use the same pool.
2. Add a configuration for the memory pool. Only works with SFTP
enabled.
3. Add the memory pool use statistics to the echoserver.
1. Remove the continue from the SFTP worker loop when there's a timeout
on the select and when there's a want_read from the wolfSSH_worker.
2. When the select has receive data ready, check it. Don't make it
conditional on want read as well.
The goal is that when there isn't data waiting on the socket, check and
process the data in the SFTP channel's buffer, always.
1. Modified SignalTcpReady() to test.h. Matched its prototype to the
other functions for TcpReady.
2. Add a timeout in WaitTcpReady() specifically for Zephyr builds.
3. Misc few cleanups.
A couple changes to keep the echoserver from spin-locking.
1. The SFTP worker should wait for data if the channel window is full.
If the client isn't sending a window adjust, it might be because of a
human time scale issue. New timeout is 60 seconds.
2. When the echoserver is waiting for something from the user and the
select times out, wait a second instead.
1. In the API test for SFTP, change the amount of the file requested to
the limit WOLFSSH_MAX_SFTP_RW.
2. The check for want read on the exit of the server worker should clear
ret if it is want read.
With the previous change for speed and timing, rekeying started having
trouble in SFTP. Each file block travels as two messages, and the
rekeying starts, but the messages are sent and lost. The client would
terminate the connection.
1. Tweak the timeout in the echoserver's SFTP loop.
2. Better checking for rekeying.
3. Returning that rekeying is happening.
wolfSSH has a reach too deep into the wolfSSH internals. SFTP should
be more like an application running on top of wolfSSH. We prefer to
keep it all in one application, so it will work well in an embedded
environment, but there needs more separation.
1. The sftp_worker loop in the echoserver should prioritize reading
the socket and processing the SSH layer messages. Next it should
ensure the transmit buffer for the SFTP channel is being written.
Last it should check the receive buffer for the SFTP for new
messages from the peer. wolfSSH_worker() will ensure data is read
from the socket, and distributed to the appropriate channel
receieve buffers, and will make sure the SSH bookkeeping is
performed.
2. SendChannelData() should also bound the send data amount by the
local maxPacketSz.
3. wolfSSH_SFTP_buffer_send() should send only one chunk of data, not
looping until everything is gone. Need to send, but cannot flood
the output.
4. Remove the call to wolfSSH_worker() from the buffer send.
5. The wolfSSH_SFTP_read() state machine should check the return
status from the call to buffer send, and if it isn't an error,
don't clear the state, allow for a want-write write again later.
6. Do not limit the file chunk size to send to WOLFSSH_MAX_SFTP_RW.
This limit is provided during SFTP negotiation. If the peer
requests that much data, send it.
6. Modify the select wrapper to have a very small microseconds value
in addition to the requested seconds.
1. Found a few cases where disabling RSA made some things either not
build or run correctly.
2. Hushed a few unused variables in gated-disabled situations.
3. Moved a temp variable closer to where it is used.
Yosuke Shimizu
aidan garske
John Safranek
hanshenrik
JacobBarthelmeh
gojimmypi
Andrew Hutchings
Anthony Hu
Juliusz Sosinowicz
Fabio Alemagna