2016-06-27T03:11:19 *** mumptai has joined #rtems 2016-06-27T03:48:11 [git] a0e49f9 by Chris Johns: [Add functions used the bridge and tap interfaces.] http://devel.rtems.org/changeset/a0e49f9/rtems-libbsd 2016-06-27T03:48:12 [git] 130fa35 by Chris Johns: [Make the domain allocation size global to allow it to set in tests. Some BSPs with multiple NIC locked in the tests due to not enough memory. This provides a simple per BSP way to increase the memory size.] http://devel.rtems.org/changeset/130fa35/rtems-libbsd 2016-06-27T03:48:12 [git] b23da31 by Chris Johns: [Change LibBSD to use shared interrupts.] http://devel.rtems.org/changeset/b23da31/rtems-libbsd 2016-06-27T03:48:12 [git] 9edacb6 by Chris Johns: [Fix spelling.] http://devel.rtems.org/changeset/9edacb6/rtems-libbsd 2016-06-27T03:48:12 [git] c6713d1 by Chris Johns: [Add DEFAULT_NETWORK_PAGE_MBUFS_SIZE to allow per BSP allocation domain sizes.] http://devel.rtems.org/changeset/c6713d1/rtems-libbsd 2016-06-27T03:48:12 [git] 55c564a by Chris Johns: [Refactor rc.conf processing to be services based.] http://devel.rtems.org/changeset/55c564a/rtems-libbsd 2016-06-27T03:48:12 [git] 7439724 by Chris Johns: [Add a services base test.] http://devel.rtems.org/changeset/7439724/rtems-libbsd 2016-06-27T03:48:13 [git] 459d67d by Chris Johns: [Move the Nexus bus driver decls available to users.] http://devel.rtems.org/changeset/459d67d/rtems-libbsd 2016-06-27T04:34:51 *** mumptai has quit IRC 2016-06-27T04:48:28 *** mumptai has joined #rtems 2016-06-27T05:30:22 *** ita|zzz is now known as ita 2016-06-27T05:36:39 *** Davidbrcz has joined #rtems 2016-06-27T06:45:28 [git] 39a650e by Chris Johns: [Add VDE and QEMU doco.] http://devel.rtems.org/changeset/39a650e/rtems-libbsd 2016-06-27T07:49:42 *** atmc has left #rtems 2016-06-27T08:21:52 *** Davidbrcz has quit IRC 2016-06-27T08:51:12 *** indy has joined #rtems 2016-06-27T09:40:20 *** Davidbrcz has joined #rtems 2016-06-27T10:18:49 *** Davidbrcz has quit IRC 2016-06-27T10:32:08 *** mumptai has quit IRC 2016-06-27T11:25:50 [git] 9835790 by Chris Johns: [Enable device sysctl.] http://devel.rtems.org/changeset/9835790/rtems-libbsd 2016-06-27T11:25:51 [git] 9362e76 by Chris Johns: [Enable the Cadence cgem device syctls.] http://devel.rtems.org/changeset/9362e76/rtems-libbsd 2016-06-27T11:27:33 [git] 591bd2d by Chris Johns: [Revert "Enable the Cadence cgem device syctls." This reverts commit 9362e7649bde3d31b6f464f3184fe527b4dcb1d2. Committed white space changes.] http://devel.rtems.org/changeset/591bd2d/rtems-libbsd 2016-06-27T11:31:15 [git] ace5fb6 by Chris Johns: [Enable the Cadence cgem device syctls.] http://devel.rtems.org/changeset/ace5fb6/rtems-libbsd 2016-06-27T12:28:11 *** hesham has quit IRC 2016-06-27T13:01:44 *** hesham has joined #rtems 2016-06-27T17:52:54 *** Davidbrcz has joined #rtems 2016-06-27T18:11:50 *** ita is now known as ita|zzz 2016-06-27T18:40:44 *** Davidbrcz has quit IRC 2016-06-27T18:58:55 *** mumptai has joined #rtems 2016-06-27T19:10:28 *** gedare has joined #rtems 2016-06-27T19:10:28 *** ChanServ sets mode: +o gedare 2016-06-27T23:40:20 *** mumptai has quit IRC 2016-06-28T03:33:05 [git] f588325 by Chris Johns: [Change the Nexus bus to defines and have the BSP declare them.] http://devel.rtems.org/changeset/f588325/rtems-libbsd 2016-06-28T03:33:06 [git] c2ec5f1 by Chris Johns: [Print the service when starting.] http://devel.rtems.org/changeset/c2ec5f1/rtems-libbsd 2016-06-28T03:33:06 [git] 90873cc by Chris Johns: [Add ftpd as a service. Add rtems-bsd-config.h for app libbsd set up.] http://devel.rtems.org/changeset/90873cc/rtems-libbsd 2016-06-28T04:24:40 *** hesham has quit IRC 2016-06-28T05:21:33 *** mumptai has joined #rtems 2016-06-28T05:23:17 *** sebhub has joined #rtems 2016-06-28T05:56:42 [trac] #2748/enhancement (v:4.10) created by sebastian.huber (Move RTEMS-specific socket wake-up to RTEMS-specific ) http://devel.rtems.org/ticket/2748 2016-06-28T05:57:45 [trac] #2748/enhancement (new) updated by chrisj (Will this break existing applications?) http://devel.rtems.org/ticket/2748 2016-06-28T06:05:54 sebhub: hi 2016-06-28T06:07:06 [trac] #2748/enhancement (new) updated by sebastian.huber (Yes, if they use the old network stack and include only .) http://devel.rtems.org/ticket/2748 2016-06-28T06:09:10 hi chris 2016-06-28T06:09:26 sebhub: just looking at adding DHCP to rc.conf support. 2016-06-28T06:09:29 there's waf 1.9.0 if you have not noticed 2016-06-28T06:09:36 *** ita|zzz is now known as ita 2016-06-28T06:09:38 Is the dhclient working? 2016-06-28T06:09:51 ita: nice. 2016-06-28T06:09:54 i don't know, I use the dhcpcd 2016-06-28T06:10:07 I see, Where did dhclient come from? 2016-06-28T06:10:13 joel 2016-06-28T06:10:26 It should be removed. The main is still main 2016-06-28T06:10:52 I am fine with removing it, I guess it was imported by accident 2016-06-28T06:11:02 Yes. It is confusing. 2016-06-28T06:11:21 I see the bridge and tap code needs devfs. 2016-06-28T06:11:50 a colleague works currently on a make_dev() implementation 2016-06-28T06:11:56 we port currently the pf firewall 2016-06-28T06:12:12 like the script parsing etc? 2016-06-28T06:12:18 for the rules? 2016-06-28T06:12:21 yes 2016-06-28T06:12:23 Nice. 2016-06-28T06:12:36 I have added a services model to rc.conf 2016-06-28T06:12:54 ftpd is an example or how to do it. 2016-06-28T06:13:00 or -> of 2016-06-28T06:13:43 I will look up dhcpd for setting up the interfaces. 2016-06-28T06:13:49 I am currently focused on the SMP stuff which I have to finish before I go on holiday 2016-06-28T06:14:16 Sure, please ping your colleague about the rc.conf services. 2016-06-28T06:14:34 I will forward your e-mails 2016-06-28T06:14:41 Great. 2016-06-28T06:15:00 Build and run rcconf02.exe for an example of how easy it is to set up 2016-06-28T06:15:26 The boot messages are now just like FreeBSD for this part of the boot. 2016-06-28T06:35:34 sebhub: why is nodhcp and nodhcpv6 placed in the dhcpcd.conf file in the tests? 2016-06-28T06:36:14 only in some tests, to test the ipv4ll 2016-06-28T06:37:04 Ah ok. 2016-06-28T06:37:36 I am still getting a valid DHCP lease. 2016-06-28T06:38:42 FreeBSD does not use dhcpcd by default 2016-06-28T06:40:20 It uses dhclient. 2016-06-28T06:46:18 sebhub: why is the dhcpcd command run in a task? 2016-06-28T07:18:37 it is a demon task 2016-06-28T07:18:57 Does it hold the command? 2016-06-28T07:19:06 command lock? 2016-06-28T07:19:07 yes, it runs forever 2016-06-28T07:19:14 no 2016-06-28T07:19:21 Ok 2016-06-28T07:19:38 if you start it twice, then i guess bad things will happen 2016-06-28T07:20:52 Should I add something to rc.conf to not allow the script to be run a second time? 2016-06-28T07:21:26 maybe dhcpd should not be a command like the others, ie not available for users to define and access 2016-06-28T07:22:06 it is an execute at most once command 2016-06-28T07:22:38 you could use a control socket to control running dhcpcd instances 2016-06-28T07:22:50 however, this doesn't work on RTEMS due to the single address space 2016-06-28T07:23:23 I think a reboot is ok if an option changes. It is typically engineered rather and user settable. 2016-06-28T07:23:26 it uses also a lock file, not sure if this is reliable on rtems 2016-06-28T07:23:42 for? 2016-06-28T07:23:57 to ensure only one instance is running 2016-06-28T07:24:35 Ah ok, I think the interface should be only available to tests and rc.conf. 2016-06-28T07:26:06 What should I do about defaultrouter? In FreeBSD is dhclient is used rc.conf hangs about waiting for a defaultouter to be set before moving on. 2016-06-28T07:26:42 I am wondering if some software needs a default route set to work. 2016-06-28T07:27:02 or if set after it does not work the same. 2016-06-28T07:27:30 hm, how is dhcpcd supported in FreeBSD? 2016-06-28T07:31:29 As a package, it uses dhclient by default 2016-06-28T07:32:25 I think I will need to poll the routing table for a valid default route and when it happens move or wait the defaultroute_delay timeout 2016-06-28T07:40:18 *** mumptai has quit IRC 2016-06-28T12:49:33 [trac] rtems-4.9.4_embed-it-part-0.patch.bz2 (Patch part 0) attached to #1728 by sebastian.huber http://devel.rtems.org/ticket/1728 2016-06-28T12:49:45 [trac] rtems-4.9.4_embed-it-part-1.patch.bz2 (Patch part 1) attached to #1728 by sebastian.huber http://devel.rtems.org/ticket/1728 2016-06-28T12:50:13 [trac] #1728/enhancement (closed) updated by sebastian.huber (Add patches received by mail for reference.) http://devel.rtems.org/ticket/1728 2016-06-28T13:18:33 [git] 9f61ad6 by Sebastian Huber: [testsuites: Increase invalid priority values Increase the invalid priority values to support test runs with alternative schedulers offering a significant higher maximum priority value, e.g. EDF.] http://devel.rtems.org/changeset/9f61ad6/rtems 2016-06-28T13:18:42 [git] 1d72f03 by Sebastian Huber: [score: Silence integer conversion warnings] http://devel.rtems.org/changeset/1d72f03/rtems 2016-06-28T13:18:52 [git] 254dc82 by Sebastian Huber: [score: Change Priority_Control to 64-bit A 32-bit Priority_Control limits the uptime to 49 days with a 1ms clock tick in case the EDF scheduler is used. Increase it to 64-bit to enable proper operation of the EDF scheduler, Close 2173.] http://devel.rtems.org/changeset/254dc82/rtems 2016-06-28T13:19:01 [git] 187a074 by Christian Mauderer: [libcsupport: Add dummy for setgroups(). The dummy for setgroups() allows applications using it to build (for example civetweb webserver).] http://devel.rtems.org/changeset/187a074/rtems 2016-06-28T13:19:11 [git] 7e6f235 by Christian Mauderer: [libnetworking: Add Move prototypes of non-portable _get*by*name/addr and _set/end*ent functions. This makes it easier to update .] http://devel.rtems.org/changeset/7e6f235/rtems 2016-06-28T13:19:21 [git] 232d6fe by Christian Mauderer: [libnetworking: Import current Import the from current FreeBSD. This allows to build some current software (e.g. libressl). Add __h_errno(). Update gethostent_r() API. Linux and FreeBSD use a common API now. Adapt the RTEMS one to provide the same one. Match gethostbyname_r() with prototype.] 2016-06-28T13:19:21 http://devel.rtems.org/changeset/232d6fe/rtems 2016-06-28T13:19:30 [git] 195d412 by Christian Mauderer: [libnetworking: Add minimal getnameinfo() This implementation just falls back to giving a string representation of the IP. It supports IPv4 only. Add test for getnameinfo().] http://devel.rtems.org/changeset/195d412/rtems 2016-06-28T13:19:40 [git] f2082cc by Christian Mauderer: [libnetworking: Import Import the from current FreeBSD. Necessary for update.] http://devel.rtems.org/changeset/f2082cc/rtems 2016-06-28T13:19:50 [git] 005c3fe by Christian Mauderer: [libnetworking: Import current Import the from current FreeBSD. This allows to build some current software (e.g. libressl).] http://devel.rtems.org/changeset/005c3fe/rtems 2016-06-28T13:19:59 [git] 657e6c9 by Christian Mauderer: [libnetworking: Import current Import the from current FreeBSD. This allows to build some current software (e.g. libressl). Add legacy support like * prototype for in_cksum(), * IPPORT_USERRESERVED, * deprecated IPCTL_RT* defines, * ip_fw_chk_t and ip_fw_ctl_t, * ip_nat_... (IP NAT hooks), 2016-06-28T13:19:59 and * IP_NAT option for get...] http://devel.rtems.org/changeset/657e6c9/rtems 2016-06-28T13:20:10 [git] 4c02385 by Christian Mauderer: [libnetworking: Import current Import the from current FreeBSD. Necessary due to changes in . Remove BSD hack from . Clean up problems with htonl(). These functions are defined in . This lead to some problems because they are defined in 2016-06-28T13:20:10 too. Add NTOHL, ... to ...] http://devel.rtems.org/changeset/4c02385/rtems 2016-06-28T13:20:19 [git] f0fc2c0 by Christian Mauderer: [libnetworking: Add This is necessary for new .] http://devel.rtems.org/changeset/f0fc2c0/rtems 2016-06-28T13:20:29 [git] 4ba4f82 by Christian Mauderer: [libnetworking: Hide SO_PRIVSTATE] http://devel.rtems.org/changeset/4ba4f82/rtems 2016-06-28T13:20:39 [trac] #2748/enhancement (closed) updated by Sebastian Huber (In [changeset:"e79a0ca75fbc87c17e220f6a80a64bff3d10c9dd/rtems"]: {{{ ...) http://devel.rtems.org/ticket/2748 2016-06-28T13:20:39 [git] e79a0ca by Sebastian Huber: [libnetworking: Move RTEMS-specific socket wake-up Close #2748.] http://devel.rtems.org/changeset/e79a0ca/rtems 2016-06-28T13:20:49 [git] 095ecbf by Christian Mauderer: [libnetworking: Import current Import the from current FreeBSD. This allows to build some current software (e.g. libressl).] http://devel.rtems.org/changeset/095ecbf/rtems 2016-06-28T13:20:58 [git] aa30e2c by Sebastian Huber: [bsp/mvme162: Include missing ] http://devel.rtems.org/changeset/aa30e2c/rtems 2016-06-28T13:25:36 *** cdesai has quit IRC 2016-06-28T15:07:53 [git] 48704cd by Mudit Jain: [arm/raspberrypi: Adding functionalities to Mailbox RPi Added functions for retrieving firmware revision, board model and board revision.] http://devel.rtems.org/changeset/48704cd/rtems 2016-06-28T15:07:53 [git] 68565b0 by Pavel Pisa: [arm/raspberrypi: run VideoCore and mailbox code through rtems.uncrustify. Some changes have been reverted to keep readability. For example squash of BCM2835_MBOX_TAG_* defines which are visually aligned in column.] http://devel.rtems.org/changeset/68565b0/rtems 2016-06-28T15:07:53 [git] b83c23e by Pavel Pisa: [arm/raspberrypi: resolve BSP warnings.] http://devel.rtems.org/changeset/b83c23e/rtems 2016-06-28T16:24:39 *** benjamint has quit IRC 2016-06-28T18:04:23 *** mumptai has joined #rtems 2016-06-28T21:20:57 *** mumptai has quit IRC 2016-06-28T22:17:57 *** gedare has quit IRC 2016-06-29T01:42:41 *** cdesai has joined #rtems 2016-06-29T05:56:46 sebhub: hi 2016-06-29T05:57:50 hi chris 2016-06-29T05:58:43 I think we need to fix the pid in libbsd. 2016-06-29T05:59:22 the process identifier? 2016-06-29T05:59:52 Yes as in BSD_DEFAULT_PID 2016-06-29T06:00:08 We need it to match getid() 2016-06-29T06:02:50 in which scenario is this important? 2016-06-29T06:04:18 To make code portable. Take a look at Steven's get route code. The most basic code to read routes needs the pid to match. The routing sockets return all to every socket and you need to filter on the pid plus other things 2016-06-29T06:04:56 The example code has "rtm->rtm_pid = pid = getpid(); " 2016-06-29T06:05:41 Can we just make calls to getpid() ie make BSD_DEFAULT_PID be getpid() 2016-06-29T06:06:08 Or make it a global set at init time 2016-06-29T06:06:19 Our pid should never change 2016-06-29T06:07:12 I am not sure why the PID in RTEMS is changeable at all 2016-06-29T06:07:35 we just have one process 2016-06-29T06:07:59 I think a global and I will set in the init and see how I go 2016-06-29T06:08:52 currently it makes no sense to me 2016-06-29T06:09:01 What does not make sense? 2016-06-29T06:09:08 Why the change? 2016-06-29T06:10:06 ok, the problem is: 2016-06-29T06:10:11 #define BSD_DEFAULT_PID 0 2016-06-29T06:10:21 pid_t getpid( void ) 2016-06-29T06:10:22 { 2016-06-29T06:10:23 Yes 2016-06-29T06:10:24 return _Objects_Local_node; 2016-06-29T06:10:25 } 2016-06-29T06:10:31 #if defined(RTEMS_MULTIPROCESSING) 2016-06-29T06:10:33 extern uint16_t _Objects_Local_node; 2016-06-29T06:10:34 #else 2016-06-29T06:10:36 #define _Objects_Local_node ((uint16_t)1) 2016-06-29T06:10:37 #endif 2016-06-29T06:10:54 so, BSD_DEFAULT_PID 1, should fix it? 2016-06-29T06:10:55 Oh. What is _Objects_Local_node? 2016-06-29T06:11:04 I will look at it. 2016-06-29T06:11:47 Why not make a global and set it to getpid()? 2016-06-29T06:11:53 Then it will not break 2016-06-29T06:12:33 its less efficient, we could add an assert(getpid() == BSD_DEFAULT_PID) in the initialization code 2016-06-29T06:16:08 I wonder if curproc->p_pid can be used and so the orginal code 2016-06-29T06:16:50 curproc is currently not used 2016-06-29T06:17:25 we have only one process in RTEMS, so we can use this fact and make PID a compile-time constant, which equals getpid() 2016-06-29T06:18:06 It is not in any hot paths. 2016-06-29T06:18:23 It is about a routing socket and polling for a route. 2016-06-29T06:18:24 *** mumptai has joined #rtems 2016-06-29T06:25:17 I will change 1 for now. I look at the struct proc in struct thread another day 2016-06-29T06:26:24 Much better, I can now read routes from the stack. 2016-06-29T06:34:31 ok, great would it be possible to add a test case for this? 2016-06-29T06:35:02 the test cases are important if we want to update to freebsd 10 or 11 some day 2016-06-29T06:35:09 test for what? 2016-06-29T06:35:21 to read the routes 2016-06-29T06:35:43 it didn't work before? 2016-06-29T06:35:48 Yes, just enable DHCP. The rc.conf requires a default route be set before moving on 2016-06-29T06:36:12 ok, so the test case in your new rc.conf stuff 2016-06-29T06:36:22 I am adding a helper routing to rtems/rtems-routes.h to get a route 2016-06-29T06:36:47 This sort of low level programming is a pain for each user to have to do. 2016-06-29T06:36:55 Yes 2016-06-29T06:38:40 I am not sure how to arrange the tests for this. Currently rcconf02 creates a test with static routes. I need another test the same with DHCP 2016-06-29T06:39:11 I can add rcconf03 the test code is small. 2016-06-29T06:39:41 Do the tests currently handle IPvLL? 2016-06-29T06:40:12 the failback IP address thing 2016-06-29T06:41:55 yes, if you connect two tests together, possibly with arphole 2016-06-29T06:42:17 see foobarserver and foobarclient 2016-06-29T06:42:28 OK. 2016-06-29T06:42:40 they are completely self organizing 2016-06-29T06:43:01 Connecting is easy with VDE 2016-06-29T06:43:07 and qemu 2016-06-29T06:44:19 The dhcpcd code is far too noisey 2016-06-29T06:44:29 I will sort this out as well. 2016-06-29T06:51:22 currently we don't filter according to syslog log level 2016-06-29T07:31:34 *** mkhoory-2 has joined #rtems 2016-06-29T07:34:51 *** mkhoory has quit IRC 2016-06-29T07:36:09 *** mkhoory-2 has quit IRC 2016-06-29T08:10:08 *** benjamint has joined #rtems 2016-06-29T08:39:06 [git] 9bb3ce3 by Sebastian Huber: [score: Fix SMP message handling According to the C11 standard only atomic read-modify-write operations guarantee that the last value written in modification order is read, see "7.17.3 Order and consistency". Thus we must use a read-modify-write in _SMP_Inter_processor_interrupt_handler() to make sure we read an up-to-date 2016-06-29T08:39:07 message.] http://devel.rtems.org/changeset/9bb3ce3/rtems 2016-06-29T08:39:07 [git] 891fa3e by Alexander Krutwig: [bsp/atsam: Add support for TCM] http://devel.rtems.org/changeset/891fa3e/rtems 2016-06-29T08:51:34 *** ganapatk has joined #rtems 2016-06-29T09:05:32 [git] 297d437 by Sebastian Huber: [score: _CPU_SMP_Send_interrupt() documentation] http://devel.rtems.org/changeset/297d437/rtems 2016-06-29T09:14:38 *** mumptai has quit IRC 2016-06-29T09:46:14 *** ganapatk has left #rtems 2016-06-29T11:19:54 *** shelf has quit IRC 2016-06-29T12:39:00 *** jkridner has joined #rtems 2016-06-29T13:51:14 *** DrJoel has joined #rtems 2016-06-29T13:51:15 *** DrJoel has joined #rtems 2016-06-29T13:51:15 *** ChanServ sets mode: +o DrJoel 2016-06-29T13:54:18 * DrJoel waves hi 2016-06-29T13:54:49 gedare is on holiday and doesn't appear to even have cell phone service. At least he didn't answer my text :) 2016-06-29T13:56:24 *** dipupo has joined #rtems 2016-06-29T13:58:48 *** benjamint has quit IRC 2016-06-29T13:59:13 *** benjamint has joined #rtems 2016-06-29T13:59:57 So who all is here and wants to go first? 2016-06-29T13:59:59 *** _sambeet has joined #rtems 2016-06-29T14:00:09 Habeeb Olufowobi 2016-06-29T14:00:17 Hi Joel 2016-06-29T14:00:20 Winner! 2016-06-29T14:00:38 Hi .. enjoying the time with Gedare away? :) 2016-06-29T14:00:48 So where are things? 2016-06-29T14:00:53 Not at all 2016-06-29T14:01:21 <_sambeet> Hi 2016-06-29T14:01:21 I haven't make significant progress since the last email i sent to the group 2016-06-29T14:01:57 So still a very basic early problem. Can you set breakpoints and step in gdb? 2016-06-29T14:02:11 I have tried to isolate the exception I have been getting but one exception keeps laeding me to another 2016-06-29T14:02:34 I have set the break at each stage of the program 2016-06-29T14:02:56 What is the first exception? Can you tell what address caused it and what type? 2016-06-29T14:02:57 but gdb keeps jumping into different files 2016-06-29T14:03:34 I am not able to isolate it in the interr.c file 2016-06-29T14:03:42 That can be normal. You may want to build with -O0 to make it easier to debug. 2016-06-29T14:04:12 But the error is somewhere else. You need to set a breakpoint at the actual assembly vector entry point and look at where it came from. 2016-06-29T14:04:20 I switched to -00 now but still the same way 2016-06-29T14:04:40 doing an objdump on the test file and mapping the exception back to a source line is the next step. 2016-06-29T14:05:15 Ok 2016-06-29T14:05:30 I will try to do objdump today 2016-06-29T14:05:54 Email devel@ asking for gdb and debugging hints for an exception. Ben, Pavel and Sebastian should be good resources. 2016-06-29T14:07:00 *** punitvara has joined #rtems 2016-06-29T14:07:12 Presently, I studying the tivaware from TI to see how they implemented some of the functions 2016-06-29T14:07:34 Ok thanks. I will reach out to them also. 2016-06-29T14:08:03 Just keep asking for help and hints. 2016-06-29T14:08:05 Who is next? 2016-06-29T14:08:16 Thanks Joel 2016-06-29T14:08:22 <_sambeet> Sir,I am facing Cmake build issues. 2016-06-29T14:08:45 <_sambeet> I am trying to compile boost using Cmake 2016-06-29T14:09:20 I hate to say this but again ask for help on devel@ or Boost's site. In this case, I think Chris Johns and likely others have built Boost for RTEMS. Ask. 2016-06-29T14:09:32 *** benjamint1 has joined #rtems 2016-06-29T14:09:45 Plus all these cross built packages need RSB recipes.. but you know that. 2016-06-29T14:09:53 *** deval has joined #rtems 2016-06-29T14:10:18 How many more packages/ 2016-06-29T14:11:08 <_sambeet> Rtt then rock base packages 2016-06-29T14:11:21 So close to the "real" project. :) 2016-06-29T14:11:23 <_sambeet> Rtt is the major component 2016-06-29T14:11:41 How many support packages are there in total? 2016-06-29T14:11:42 *** benjamint has quit IRC 2016-06-29T14:11:42 *** benjamint1 is now known as benjamint 2016-06-29T14:12:47 <_sambeet> All support packages work fine ....only rock packages are left 2016-06-29T14:13:03 <_sambeet> But cmake is confusing 2016-06-29T14:13:18 <_sambeet> So it's taking time 2016-06-29T14:13:32 I agree. Ask for help. There should be someone in the community who can answer questions 2016-06-29T14:13:36 Anything else? 2016-06-29T14:13:40 Who's next? 2016-06-29T14:13:54 Should I start ? 2016-06-29T14:14:31 I have already completed PWM driver and tested successfully. I have already sent patches as well 2016-06-29T14:14:49 <_sambeet> Yes sir I will send my queries in users list as per gedare's suggestions 2016-06-29T14:15:02 <_sambeet> Thank you sir 2016-06-29T14:15:20 Yesterday My mentor just asked me to improve one function. I guess within 2 days we can wrap up PWM and merge into RTEMS 2016-06-29T14:15:37 (hi all, punitvara) 2016-06-29T14:15:50 *** _sambeet has quit IRC 2016-06-29T14:16:05 beng-nl... take over for a minute. Guy is outside working on the house and has question 2016-06-29T14:16:18 *** mudit1729 has joined #rtems 2016-06-29T14:16:52 Hi Ben .. I have just said about polishing pwm code 2016-06-29T14:23:38 DrJoel: ok 2016-06-29T14:24:15 punitvara: are you stuck on anything in PWM? 2016-06-29T14:25:13 No I have finished polishing as Martin told. Need to test one more time because I changed pinmux function 2016-06-29T14:25:25 you have made all changes that martin requested? 2016-06-29T14:25:52 and not forgotten to test that your build does not break the beagleboard xm build, and that your code does not try to touch the hardware if built in beagleboard xm mode :-) ? 2016-06-29T14:26:16 Yes I made all the changes as Martin suggested 2016-06-29T14:27:00 I understood correctly I need to build BB xm BSP and check whether I am able to build it successful or not right ? 2016-06-29T14:27:12 *If 2016-06-29T14:27:37 yes, (a) the build must work and (b) a bbxm application should not be able to use the PWM functions because the hardware is (presumably) different 2016-06-29T14:27:45 that should cause either a link or runtime error 2016-06-29T14:28:09 punitvara: will you submit a new patch soon? 2016-06-29T14:28:29 or new patches. 2016-06-29T14:28:35 anything else? 2016-06-29T14:28:36 Within a hour or two. I have just built it 2016-06-29T14:28:39 punitvara: good 2016-06-29T14:28:40 anything else? 2016-06-29T14:28:58 Going to test it that it. And we are going to work on I2C Right ? 2016-06-29T14:29:03 seems so. 2016-06-29T14:29:04 after pwm 2016-06-29T14:29:16 ok 2016-06-29T14:29:18 punitvara: and pester your mentor if you have to, it's his job to help you too ;-) 2016-06-29T14:29:21 ok good. 2016-06-29T14:29:23 who's next? 2016-06-29T14:30:06 any other GSOC student here? 2016-06-29T14:30:09 there should be. 2016-06-29T14:30:14 Can I go next ? 2016-06-29T14:30:24 deval: yes please. 2016-06-29T14:30:31 Currently I am trying to port usb_ethernet driver. The work is under progress. Resolving errors while compiling. 2016-06-29T14:30:55 beng-nl: good you stepped in there. That as a project you were better for. Mr ARM Beagle. :) 2016-06-29T14:30:59 I am not stuck at any point 2016-06-29T14:31:11 deval: Do you have much that isn't merged? 2016-06-29T14:31:22 deval: it seems DrJoel is taking over again :) 2016-06-29T14:31:31 DrJoel: no problem, i got lucky i knew about the current project ;-) 2016-06-29T14:31:59 Yes I have ported bcm283x_dwcotg driver 2016-06-29T14:32:07 which is not yet merged 2016-06-29T14:32:12 to all GSOC students: work proof submission needs to focus on merged work with URLs, tickets, etc. If not merged, then available in a permanent way and documented so someone can pick it up. Hopefully you will want to finish it though. :) 2016-06-29T14:32:25 beng-nl: You are also very qualified on this one. :) 2016-06-29T14:32:36 and you have some USB device working? 2016-06-29T14:33:56 So the usb from the processor is connected to LAN9514, Currently I am porting the driver for that 2016-06-29T14:34:08 Talk to sebhub but it may be worth merging base FreeBSD code and later patches, rtems specific stuff 2016-06-29T14:34:24 after that I will be able to demo any working USB device 2016-06-29T14:34:27 great. Who made that part? 2016-06-29T14:34:59 I did not get you 2016-06-29T14:35:11 Who manufactured LAN9514? 2016-06-29T14:35:31 microchip 2016-06-29T14:35:34 Ahhh 2016-06-29T14:36:06 We have drivers for that in FreeBSD 2016-06-29T14:36:10 Not a necessity but it is possible and might be easier to debug USB only part with a mouse. It should be quiet until you move it. 2016-06-29T14:36:32 That greatly simplifies things but it is still a challenge. :) 2016-06-29T14:36:55 So your current plan is to debug LAN+USB together? 2016-06-29T14:37:27 LAN9514 is a USB+LAN, I will not be able to interface USB without adding this driver 2016-06-29T14:37:45 OK. Then that's the path forward. 2016-06-29T14:37:58 yeah 2016-06-29T14:38:05 You said no issues.. so just keep plugging and report on progress. 2016-06-29T14:38:13 if nothing else, who is next? 2016-06-29T14:38:14 Sure 2016-06-29T14:38:19 Thank you sir 2016-06-29T14:38:26 No problem. :) 2016-06-29T14:39:09 darnir: are you there? Feeling any better? 2016-06-29T14:39:44 dcc chat beng-nl 2016-06-29T14:39:53 well that didn't work 2016-06-29T14:39:57 anyone awake? 2016-06-29T14:39:59 Can I go next ? 2016-06-29T14:40:03 Sure.. :) 2016-06-29T14:40:33 I am working on porting the SD card driver and the imporving the mailbox functionality on rpi 2016-06-29T14:40:53 *** benjamint1 has joined #rtems 2016-06-29T14:41:16 The mailbox code that was pushed was merged by Pavel and some other changes suggested by Gedare are also completed. 2016-06-29T14:41:52 *** benjamint has quit IRC 2016-06-29T14:41:52 *** benjamint1 is now known as benjamint 2016-06-29T14:41:58 Now the SD card driver, the SDHCI controller, DMA controller, MMC/SD bus gets initialized, however the SD card is not detected. 2016-06-29T14:42:01 Awesome! I saw it looked like a breakthrough had happened. Is this considered a permanent fix for the caching or is that ongoing? 2016-06-29T14:42:25 That sounds like big progress! 2016-06-29T14:42:42 Is this just something you will have to debug? Any blockers? 2016-06-29T14:42:44 It presently won't be hindering development, for RPi peripherals if a particular function us ysed 2016-06-29T14:43:13 Yes, I will have to debug the device tree to understand why the card is not getting detected 2016-06-29T14:44:05 I will be pushing the code for review on devel. 2016-06-29T14:44:07 That can be tricky. Just ask on devel@ for help. kiwichris and sebhub are good at this stuff. 2016-06-29T14:44:15 Marvelous on posting the code. 2016-06-29T14:44:28 Anything you need? Anything else? 2016-06-29T14:44:31 Yes. I will be doing that by tomorrow 2016-06-29T14:44:51 I required a discussion regarding the next items that are in my deliverables list 2016-06-29T14:45:15 They include DMA support for I2C and SPI and adding support for second SPI device. 2016-06-29T14:45:44 One of the topics of discussion is which peripherable will be used to test the changes 2016-06-29T14:46:08 Are any built onto the Pi? 2016-06-29T14:46:18 brb going to get coffiee 2016-06-29T14:47:11 I will have to do a bit of reading before I can comment on that. 2016-06-29T14:49:06 ok. please discuss on the list. plenty of people will have ideas. Alan Cudmore should be one of them. He had a cool PiHat board on his last year at a workshop. Was walking around running NASA Core Flight Software on it. :) 2016-06-29T14:49:14 Any blockers? 2016-06-29T14:49:50 I will be asking all this all devel. 2016-06-29T14:50:02 Are you going to get the SD card detection solved before moving on or working both in parallel? 2016-06-29T14:50:42 I will be pushing the code and will be asking for review. 2016-06-29T14:50:52 *** benjamint1 has joined #rtems 2016-06-29T14:50:56 In the mean time I can utilise my time by working on the next deleverable 2016-06-29T14:52:15 OK. Sounds like we can move on.. 2016-06-29T14:52:18 who's next? 2016-06-29T14:53:09 *** benjamint has quit IRC 2016-06-29T14:53:09 *** benjamint1 is now known as benjamint 2016-06-29T14:53:53 ping.. anyone left? 2016-06-29T14:54:31 darnir? 2016-06-29T14:54:45 Wonjun__? 2016-06-29T14:55:33 well unless they speak up, that's it. 2016-06-29T14:55:45 anyone have anything else? 2016-06-29T14:56:23 beng-nl: Michele and I theorectically have 1h 20m at Schilpol next Friday. I am sorry it isn't longer and we could see you 2016-06-29T14:56:35 *** deval has quit IRC 2016-06-29T14:57:21 *** mudit1729 has quit IRC 2016-06-29T14:58:44 well.. I need to elocate to the office.. I can't spend all day in my PJs at home. :) 2016-06-29T14:58:57 Please ask questions and don't get blocked 2016-06-29T14:59:05 *** DrJoel has quit IRC 2016-06-29T15:08:49 *** punitvara has quit IRC 2016-06-29T15:34:42 *** ketul has joined #rtems 2016-06-29T15:39:07 *** ketul has quit IRC 2016-06-29T18:06:29 *** mumptai has joined #rtems 2016-06-29T18:38:08 DrJoel: Hi. I sent you an email earlier today stating that I wouldn't be able to make it. Doctor had called me in for a checkup at the same time as the meeting. 2016-06-29T18:38:26 Getting better, which means I am starting to get more time to work. 2016-06-29T19:03:09 *** stryx` has quit IRC 2016-06-29T19:05:11 *** stryx` has joined #rtems 2016-06-29T22:57:06 *** mumptai has quit IRC 2016-06-29T23:20:15 [git] 4a2b844 by Chris Johns: [Default syslog priority to INFO. Add rtems_setlogpriority to configure the priority. The rc.conf has 'syslog_priority' where the priority is the name, eg syslog_priority="debug" sets the priority to "debug".] http://devel.rtems.org/changeset/4a2b844/rtems-libbsd 2016-06-29T23:20:15 [git] 45960a3 by Chris Johns: [Add rtems_get_route to fetch a specific route from the routing tables.] http://devel.rtems.org/changeset/45960a3/rtems-libbsd 2016-06-29T23:20:15 [git] c980a04 by Chris Johns: [Fix fomatting and C++ guard.] http://devel.rtems.org/changeset/c980a04/rtems-libbsd 2016-06-29T23:20:15 [git] 912b685 by Chris Johns: [Make the LibBSD PID match getpid().] http://devel.rtems.org/changeset/912b685/rtems-libbsd 2016-06-29T23:20:15 [git] 573b198 by Chris Johns: [Add DHCP support to rc.conf.] http://devel.rtems.org/changeset/573b198/rtems-libbsd 2016-06-29T23:25:21 [git] dc5b267 by Chris Johns: [Fix wording and spelling.] http://devel.rtems.org/changeset/dc5b267/rtems-libbsd 2016-06-30T01:43:39 [git] f01b1f7 by Chris Johns: [Add RTEMS_BSD_CONFIG_DOMAIN_PAGE_MBUFS_SIZE. The domain allocator page abd mbuf size can be configured by the user using RTEMS_BSD_CONFIG_DOMAIN_PAGE_MBUFS_SIZE.] http://devel.rtems.org/changeset/f01b1f7/rtems-libbsd 2016-06-30T05:58:59 [git] b6b25a9 by Sebastian Huber: [score: Fix thread lock on SMP configurations] http://devel.rtems.org/changeset/b6b25a9/rtems 2016-06-30T05:58:59 [git] 0298772 by Sebastian Huber: [score: Avoid atomic fences for thread wait flags The use of atomic fences is brittle and may break due to changes in different areas which is hard to manage.] http://devel.rtems.org/changeset/0298772/rtems 2016-06-30T07:55:23 [git] 09c87fb by Sebastian Huber: [score: Strong APA scheduler processor allocation Use _Scheduler_SMP_Allocate_processor_exact() to prevent unexpected migrations.] http://devel.rtems.org/changeset/09c87fb/rtems 2016-06-30T07:55:23 [git] 729cf69 by Sebastian Huber: [bsps: Do not use fast idle clock for SMP We may have more than one idle thread. Thus, the clock driver fast idle logic would be a bit more complicated.] http://devel.rtems.org/changeset/729cf69/rtems 2016-06-30T07:55:23 [git] 4fe0e97 by Sebastian Huber: [smptests/smpstrongapa01: Add test cases] http://devel.rtems.org/changeset/4fe0e97/rtems 2016-06-30T09:01:28 [git] 4280dff by Sebastian Huber: [smptests/smpstrongapa01: Simplify] http://devel.rtems.org/changeset/4280dff/rtems 2016-06-30T11:12:19 [trac] #2749/defect (v:4.11) created by sebastian.huber (rtems_task_set_scheduler() has insufficient parameters) http://devel.rtems.org/ticket/2749 2016-06-30T11:35:31 [trac] #2749/defect (new) updated by sebastian.huber (empty comment) http://devel.rtems.org/ticket/2749 2016-06-30T14:00:49 [trac] #2750/defect (v:4.12) created by joel.sherrill (Compile Error When Multiprocessing Enabled) http://devel.rtems.org/ticket/2750 2016-06-30T17:21:22 *** dougk_ has quit IRC 2016-06-30T17:33:17 *** dougk_ has joined #rtems 2016-07-01T05:50:48 [git] fc26479 by Chris Johns: [Add ftpd initailize comment.] http://devel.rtems.org/changeset/fc26479/rtems-libbsd 2016-07-01T05:50:48 [git] 65c65bb by Chris Johns: [Add telnetd as service to rc.conf.] http://devel.rtems.org/changeset/65c65bb/rtems-libbsd 2016-07-01T06:36:24 [git] 4dec435 by Chris Johns: [Add a wait for key press to test rccof02.] http://devel.rtems.org/changeset/4dec435/rtems-libbsd 2016-07-01T06:44:09 [git] 34f6271 by Chris Johns: [Provide empty RTEMS_BSD_CFGDECL_.* defines if not used.] http://devel.rtems.org/changeset/34f6271/rtems-libbsd 2016-07-01T07:42:38 [trac] #2751/defect (v:4.11) created by sebastian.huber (Thread dispatch via interrupt is broken at least on ARM and PowerPC) http://devel.rtems.org/ticket/2751 2016-07-01T07:58:41 [trac] GSoC/2016 edited by hwj1593 (empty comment) http://devel.rtems.org/wiki/GSoC/2016 2016-07-01T08:16:08 [trac] GSoC/2016 edited by hwj1593 (empty comment) http://devel.rtems.org/wiki/GSoC/2016 2016-07-01T09:57:39 [trac] #2749/defect (new) updated by sebastian.huber (empty comment) http://devel.rtems.org/ticket/2749 2016-07-01T09:58:26 [git] 4142f73 by Sebastian Huber: [score: Fix priority affinity SMP scheduler Bug introduced by 9bfad8cd519f17cbb26a672868169fcd304d5bd5.] http://devel.rtems.org/changeset/4142f73/rtems 2016-07-01T09:58:36 [git] 23dedc4 by Sebastian Huber: [score: Next try to fix thread lock on SMP] http://devel.rtems.org/changeset/23dedc4/rtems 2016-07-01T09:58:45 [trac] #2751/defect (new) updated by Sebastian Huber (In [changeset:"8d5b03802e99e581c360e9a2cf67856596ec824c/rtems"]: {{{ ...) http://devel.rtems.org/ticket/2751 2016-07-01T09:58:46 [git] 8d5b038 by Sebastian Huber: [score: Workaround for #2751 The ARM and PowerPC interrupt epilogues call _Thread_Dispatch() with interrupts disabled (counter example: SPARC). On SMP configurations, since inter-processor interrupts set the thread dispatch necessary indicator this prevents a thread dispatch notification in post-switch handlers (which all run 2016-07-01T09:58:46 with interrupts disabled)...] http://devel.rtems.org/changeset/8d5b038/rtems 2016-07-01T09:58:55 [trac] #2749/defect (closed) updated by Sebastian Huber (In [changeset:"c0bd0064ac41f0602c0abfe494dbe140d7c5282f/rtems"]: {{{ ...) http://devel.rtems.org/ticket/2749 2016-07-01T09:58:55 [git] c0bd006 by Sebastian Huber: [rtems: Fix rtems_task_set_scheduler() API Task priorities are only valid within a scheduler instance. The rtems_task_set_scheduler() directive moves a task from one scheduler instance to another using the current priority of the thread. However, the current task priority of the source scheduler instance is undefined in the 2016-07-01T09:58:56 target scheduler instance. A...] http://devel.rtems.org/changeset/c0bd006/rtems 2016-07-01T09:59:05 [trac] #2750/defect (closed) updated by Sebastian Huber (In [changeset:"4cb13c399827a39a9108a631e6afddf6d96de60c/rtems"]: {{{ ...) http://devel.rtems.org/ticket/2750 2016-07-01T09:59:06 [git] 4cb13c3 by Sebastian Huber: [score: Fix MPCI message layout Restore the 32-bit priority field in MP_packet_Prefix. Bug introduced by 254dc82daf8cbd6922376fcbb81c31e21cbf4d16. Close #2750.] http://devel.rtems.org/changeset/4cb13c3/rtems 2016-07-01T10:01:25 [trac] #2695/defect (closed) updated by sebastian.huber (Fixed in this RSB version: ...) http://devel.rtems.org/ticket/2695 2016-07-01T10:02:34 [trac] #2737/defect (new) updated by sebastian.huber ([rtems/b61d5cac7c5f1ba801a8d0f896313b2e5cd01111]) http://devel.rtems.org/ticket/2737 2016-07-01T13:17:48 [git] df2177a by Sebastian Huber: [score: Change scheduler node init and destroy Provide the scheduler node to initialize or destroy to the corresponding operations. This makes it possible to have more than one scheduler node per thread.] http://devel.rtems.org/changeset/df2177a/rtems 2016-07-01T13:17:49 [git] 2e3ba71 by Sebastian Huber: [net: Fix byte order issue for getnameinfo()] http://devel.rtems.org/changeset/2e3ba71/rtems 2016-07-01T13:55:15 *** sebhub has quit IRC 2016-07-01T17:04:40 *** stryx` has quit IRC 2016-07-01T17:05:44 *** stryx` has joined #rtems 2016-07-01T19:41:53 *** edwardk has quit IRC 2016-07-01T19:42:28 *** dipupo has quit IRC 2016-07-01T19:42:54 *** modwizcode has quit IRC 2016-07-01T19:44:21 *** modwizcode has joined #rtems 2016-07-01T19:44:36 *** edwardk has joined #rtems 2016-07-01T19:55:04 *** dipupo has joined #rtems 2016-07-02T11:15:51 *** Davidbrcz has joined #rtems 2016-07-02T11:31:04 *** Davidbrcz has quit IRC 2016-07-02T16:16:56 *** verm__ has quit IRC 2016-07-02T16:17:36 *** verm__ has joined #rtems 2016-07-02T17:30:58 *** cdesai has quit IRC 2016-07-02T17:33:18 *** cdesai has joined #rtems 2016-07-02T23:29:53 *** benjamint has quit IRC 2016-07-02T23:30:08 *** benjamint has joined #rtems 2016-07-03T15:57:52 *** mumptai has joined #rtems 2016-07-03T21:09:00 *** dipupo has quit IRC 2016-07-03T21:27:22 *** Davidbrcz has joined #rtems 2016-07-03T21:56:52 *** Davidbrcz has quit IRC 2016-07-03T23:20:44 *** mumptai has quit IRC