This patch adds epoll option for the eloop implementation. This can be
selected with the CONFIG_ELOOP_EPOLL=y build option.
[merit]
See Table1.
Table1. comparison table
+--------+--------+-----------+------------+-------------+
| | add fd | remove fd | prepare fd | dispatch fd |
+--------+--------+-----------+------------+-------------+
| select | O(1) | O(1) | O(N) | O(N) |
+--------+--------+-----------+------------+-------------+
| poll | O(1) | O(1) | O(N) | O(N) |
+--------+--------+-----------+------------+-------------+
| epoll | O(1) | O(1) | 0 | O(M) |
+--------+--------+-----------+------------+-------------+
"add fd" is addition of fd by eloop_sock_table_add_sock().
"remove fd" is removal of fd by eloop_sock_table_remove_sock().
"prepare fd" is preparation of fds before wait in eloop_run().
"dispatch fd" is dispatchment of fds by eloop_sock_table_dispatch().
"N" is all watching fds.
"M" is fds which could be dispatched after waiting.
As shown in Table1, epoll option has better performance on "prepare fd" column.
Because select/poll option requires setting fds before every select()/poll().
But epoll_wait() doesn't need it.
And epoll option has also better performance on "dispatch fd" column.
Because select/poll option needs to check all registered fds to find out
dispatchable fds. But epoll option doesn't require checking all registered fds.
Because epoll_wait() returns dispatchable fd set.
So epoll option is effective for GO/AP functionality.
[demerit]
The epoll option requires additional heap memory. In case of P2P GO, it is
about 8K bytes.
Signed-off-by: Masashi Honma <masashi.honma@gmail.com>
scard_gsm_auth() used SIM_CMD_GET_RESPONSE for both SIM and USIM. Convert
the command into USIM_CMD_GET_RESPONSE for USIM.
Since commit eb32460029 ("Fix switching from EAP-SIM to EAP-AKA/AKA'")
EAP-SIM is using the USIM if available. This triggers a probably ancient
bug in scard_gsm_auth(), which results in sending the wrong get response
command to the USIM. Thus, EAP-SIM stopped to work after this change on
USIMs that expect the proper command.
Signed-off-by: Simon Baatz <gmbnomis@gmail.com>
It looks like the demangle.h from binutils-dev is not installed that
commonly anymore. Since we need only two defines from that file, replace
the header file with those defines to make it easier to build with
WPA_TRACE_BFD=y.
Signed-off-by: Jouni Malinen <j@w1.fi>
These functions are not standard and do not exist in all systems, e.g.,
variants of Android. Instead use the macros defined in common.h.
Signed-off-by: Ilan Peer <ilan.peer@intel.com>
Update the radiotap parser to the latest version of the
http://git.sipsolutions.net/radiotap.git/ library to get
parsing for vendor namespaces.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
wpa_supplicant and wpa_cli had already moved to allowing up to 4096 byte
buffer size to be used for control interface commands. This was limited
by the line edit buffer in interactive mode. Increase that limit to
match the other buffers to avoid artificially truncating long commands.
Signed-off-by: Jouni Malinen <j@w1.fi>
Do not truncate CURLINFO entries on first linefeed to get full IN/OUT
headers and data into debug log. Use wpa_hexdump_ascii() if any
non-displayable characters are included. Remove the separate header/data
debug dumps since all that information is now available from the debug
callback.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
This hs20-osu-client client command can be used to parse a DER encoded
X.509v3 certificate with the logotype extensions and
id-wfa-hotspot-friendlyName values shown in detail.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
bfd_demangle() call could be skipped if data.function == NULL. Make sure
the already freed aname pointer cannot be used again in such a case.
Signed-off-by: Jouni Malinen <j@w1.fi>
Commit e2f5a9889a was supposed to prevent
new scan request from pushing out the old one. However, it did not
really do that since eloop_deplete_timeout() returned 0 both for the
case where the old timeout existed (and was sooner) and if the old
timeout did not exist. It returned 1 only for the case where an old
timeout did exist and was larger than the new requested value. That case
used to result in wpa_supplicant_req_scan() rescheduling the timeout,
but hew code in eloop_deplete_timeout() did the exact same thing and as
such, did not really change anything apart from the debug log message.
Extend the eloop_deplete_timeout() (and eloop_replenish_timeout() for
that matter since it is very similar) to return three different values
based on whether the timeout existed or not and if yes, whether it was
modified. This allows wpa_supplicant_req_scan() to schedule a new
timeout only in the case there was no old timeout.
Signed-hostap: Jouni Malinen <j@w1.fi>
These were somewhat more hidden to avoid direct use, but there are now
numerous places where these are needed and more justification to make
the extern int declarations available from wpa_debug.h. In addition,
this avoids some warnings from sparse.
Signed-hostap: Jouni Malinen <j@w1.fi>
This can be used to silence compiler warnings in cases where #ifdef
blocks can leave some variables or functions unused and there is no
cleaner way of avoiding the warnings.
Signed-hostap: Jouni Malinen <j@w1.fi>
Just the usual, with a new function os_reltime_initialized()
thrown in that checks whether time has ever been retrieved
(time can't be completely zero).
Signed-hostap: Johannes Berg <johannes.berg@intel.com>
This helper functions checks whether a given entry has expired,
given the last active timestamp, the current time, and a timeout.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Relative time shouldn't be calculated based on gettimeofday
because that clock can jump (e.g., when the time is adjusted
by the system administrator.)
On systems where that is available, use CLOCK_BOOTTIME (on
fairly recent Linux systems, this clock takes into account
the time spend suspended) or CLOCK_MONOTONIC (on Linux and
some POSIX systems, this clock is just freely running with
no adjustments.)
Reported-by: Holger Schurig <holgerschurig@gmail.com>
Signed-hostap: Johannes Berg <johannes.berg@intel.com>
This allows wlantest debug log output to be directed to a file so that
RELOG command can be used to rotate files more easily than stdout.
Signed-hostap: Jouni Malinen <j@w1.fi>
This was supposed to be a minimal sample of eloop wrapper, but it is
unclear whether this is of that much use and the file has not been kept
up-to-date. Remove this file to reduce maintenance effort. The other
eloop*.c files can be used as a starting point if something new is
needed.
Signed-hostap: Jouni Malinen <j@w1.fi>
Commit eb32460029 left an unneeded
sim_type argument to scard_init(). Remove that unnecessary argument to
clean up the implementation.
Signed-hostap: Masashi Honma <masashi.honma@gmail.com>
eloop_replenish_timeout() finds a registered matching
<handler,eloop_data,user_data> timeout. If found, replenishes
the timeout if remaining time is less than the requested time.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
os_strlpcy() should be used instead of os_strncpy() to guarantee null
termination. Since there are no remaining strncpy uses, remove
os_strncpy() definition.
Signed-hostap: Jouni Malinen <j@w1.fi>
printf_decode() fills in a binary buffer and returns the length of
the written data. This did not use null termination since initial
use cases used the output as a binary value. However, Hotspot 2.0
cred block values are also using this for parsing strings. Those
cases could end up without proper null termination depending on what
os_malloc() ends up getting as the memory buffer. Fix these and make
printf_decode() more convenient by forcing the output buffer to be
null terminated.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
Compiling hostapd with VHT enabled on a big endian machine resulted in
an undefined symbol error. Fix this by defining the missing macro.
Signed-hostap: Michal Kazior <michal.kazior@tieto.com>
Commit 3962b65858 removed the Xcode
project files, but missed the defines in build_config.h. Remove these
since there are no users for them in the current snapshot.
Signed-hostap: Jouni Malinen <j@w1.fi>
These events are sent as a special case to both the group interface and
"parent interface" (i.e., the interface that was used for managing P2P
negotiation). The latter is not really correct event, so get rid of it
with the new global control interface design where there is no need to
support legacy upper layer implementations.
Signed-hostap: Jouni Malinen <j@w1.fi>
This function can be used instead of wpa_msg() and wpa_msg_ctrl() to
indicate that an event is not specific to a network interface.
Signed-hostap: Jouni Malinen <j@w1.fi>
The HIST_ENTRY and its variables are allocated within libreadline, so
they won't have the WPA_TRACE special header and cannot be freed with
os_free(). Use free() to avoid issues during wpa_cli termination if any
of the new commands added to the history are to be removed (e.g.,
set_network could include a password).
Signed-hostap: Jouni Malinen <j@w1.fi>
DFS implementation requires to run an eventloop while monitoring
the Channel Availability Check (CAC). After that, the "real" event
loop is started, and should not fail doing so.
Signed-hostap: Simon Wunderlich <siwu@hrz.tu-chemnitz.de>
currbuf_valid needs to be cleared when an old command from history is
processed to avoid leaving a bogus entry that makes history_prev() skip
the last entry in history.
Signed-hostap: Jouni Malinen <j@w1.fi>
There are quite a few places in the current implementation where a nul
terminated string is generated from binary data. Add a helper function
to simplify the code a bit.
Signed-hostap: Jouni Malinen <j@w1.fi>
This option can be used only for global parameters that are not going
to be changed from settings.
Signed-off-by: Dmitry Shmidt <dimitrysh@google.com>
Signed-off-by: Iliyan Malchev <malchev@google.com>
ANDROID_SETGROUPS_OVERRIDE macro can now be used to override setgroups()
values based on build configuration.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
These verify that a proper pointer type is used and in addition, seems
to get rid of some false static analyzer warnings.
Signed-hostap: Jouni Malinen <j@w1.fi>
Use an explicit pointer to the beginning of the buffer and a flag
to indicate whether that is to external data or not. This avoids
a branch whenever accessing the buffer and helps some static
analyzers to understand the wpabuf memory uses better.
Signed-hostap: Jouni Malinen <j@w1.fi>
This function can be used as a wrapper for os_realloc(ptr, nmemb * size)
when a reallocation is used for an array. The main benefit over
os_realloc() is in having an extra check to catch integer overflows in
multiplication. This is similar to the os_zalloc() to os_calloc() move.
Signed-hostap: Jouni Malinen <j@w1.fi>
This function can be used as a wrapper for os_zalloc(nmemb * size) when
an allocation is used for an array. The main benefit over os_zalloc() is
in having an extra check to catch integer overflows in multiplication.
Signed-hostap: Jouni Malinen <j@w1.fi>
The new ssid2 parameter can be used as an alternative mechanism for
configuring SSID for hostapd. It uses the same formats that
wpa_supplicant uses in the configuration file for strings.
Signed-hostap: Jouni Malinen <j@w1.fi>
Instead of masking out non-ASCII characters with underscores, espace the
SSID data using rules compatible with printf.
Signed-hostap: Jouni Malinen <j@w1.fi>
This can be used to build ASCII strings from binary data that is
more likely to use ASCII (i.e., text format is more natural
option than hexdump, but there is possibility of some non-ASCII
characters).
Signed-hostap: Jouni Malinen <j@w1.fi>
This new mechanism can be used to make wpa_supplicant using external
storage (e.g., key store in the operating system) for passwords,
passphrases, and PSKs. This commit is only adding the framework part
needed to support this, i.e., no actual configuration parameter can
yet use this new mechanism. In addition, only a simple test backend
is added to allow developer testing of the functionality.
Signed-hostap: Jouni Malinen <j@w1.fi>
Without this, we can get into a tight loop because the
code in general doesn't add eloop exception handlers,
so socket reporting the POLLERR would never be read.
With this change, any socket with POLLERR or POLLHUP
asserted will be handled by the read logic.
Signed-hostap: Ben Greear <greearb@candelatech.com>
This was never really completed and is of not much use at this point,
so clean up the repository by removing the Symbian-specific changes.
Signed-hostap: Jouni Malinen <j@w1.fi>
Add the option (-T) to wpa_supplicant to log all debug messages into the
kernel tracing, allowing to aggregate kernel debugging with
wpa_supplicant debugging and recording all with trace-cmd.
Since tracing has relatively low overhead and can be filtered
afterwards, record all messages regardless of log level. However, it
will honor the -K option and not record key material by default.
Signed-hostap: Johannes Berg <johannes.berg@intel.com>
This DO has a maximum length of 127 bytes (TS 102 221), so better use
large enough buffer when reading the record.
Signed-hostap: Jouni Malinen <j@w1.fi>
New global configuration parameters pcsc_reader and pcsc_pin can now be
used to initialize PC/SC reader context at start of wpa_supplicant.
Signed-hostap: Jouni Malinen <j@w1.fi>
eloop_sock_table_add_sock() needs to fail if pollfd array allocation
fails instead of returning success and leaving behind no buffer.
Signed-hostap: Jouni Malinen <j@w1.fi>
When using more than around 200 virtual stations, we start hitting the
max number of file descriptors supported by select(). This patch adds
support for poll(), which has no hard upper limit.
Signed-hostap: Ben Greear <greearb@candelatech.com>
This patch allows to log hexdumps into syslog.
This is useful when testing, as syslog's network logging
helps to collect debug outputs from several machines.
Signed-hostapd: Nicolas Cavallari <cavallar@lri.fr>
The EF-AD (administrative data) file may contain information about the
length of the MNC (2 or 3 digits) in the IMSI. This can be used to
construct the realm according to 3GPP TS 23.003 during EAP-SIM or
EAP-AKA authentication.
Signed-hostap: Simon Baatz <gmbnomis@gmail.com>
It looks like some USIM cards respond with 0x67 (Wrong length) instead
of 0x6c to 00 b2 01 04 ff. This was getting rejected in
scard_get_record_len(). ETSI TS 102 221 is not very clear on this
detail, but it looks fine to accept the 0x67 error value, too, to learn
the record length.
Signed-hostap: Jouni Malinen <j@w1.fi>
