From bugmaster at FreeBSD.org Mon Sep 1 11:06:53 2008 From: bugmaster at FreeBSD.org (FreeBSD bugmaster) Date: Mon Sep 1 11:07:28 2008 Subject: Current problem reports assigned to freebsd-embedded@FreeBSD.org Message-ID: <200809011106.m81B6qTk068381@freefall.freebsd.org> Current FreeBSD problem reports Critical problems Serious problems Non-critical problems S Tracker Resp. Description -------------------------------------------------------------------------------- o misc/15876 embedded [picobsd] PicoBSD message of the day problems o kern/42728 embedded [picobsd] many problems in src/usr.sbin/ppp/* after c o misc/52256 embedded [picobsd] picobsd build script does not read in user/s o kern/101228 embedded [nanobsd] [patch] Two more entries for FlashDevice.sub 4 problems total. From info at martenvijn.nl Sat Sep 6 09:26:04 2008 From: info at martenvijn.nl (Marten Vijn) Date: Sat Sep 6 09:26:15 2008 Subject: Tinybsd info and pictures Message-ID: <1220692251.5919.20.camel@mvn-desktop> Hi I am preparing an presentation/workshop about tinybsd. I would like to know who using Tinybsd and how. If you use Tinybsd could you answer some questions and send me offlist some pictures. Only send it if you allow me to you use this for presentation. - What is your name or organisation (anonymous is ok) - Where are you using Tinybsd (Country)? - What is the purpose? - How long do you use it? - How many installations do you run? - What hardware do you use? - Which version of FreeBSD is used? Only send this if you allow me to you use this for presentation/workshop. (slides online) kind regards, Marten -- Marten Vijn http://martenvijn.nl http://opencommunitycamp.org http://wifisoft.org From bugmaster at FreeBSD.org Mon Sep 8 02:22:18 2008 From: bugmaster at FreeBSD.org (FreeBSD bugmaster) Date: Mon Sep 8 02:22:43 2008 Subject: Current problem reports assigned to freebsd-embedded@FreeBSD.org Message-ID: <200809080222.m882MH8B006628@freefall.freebsd.org> The following is a listing of current problems submitted by FreeBSD users. These represent problem reports covering all versions including experimental development code and obsolete releases. S Tracker Resp. Description -------------------------------------------------------------------------------- o kern/101228 embedded [nanobsd] [patch] Two more entries for FlashDevice.sub o misc/52256 embedded [picobsd] picobsd build script does not read in user/s o kern/42728 embedded [picobsd] many problems in src/usr.sbin/ppp/* after c o misc/15876 embedded [picobsd] PicoBSD message of the day problems 4 problems total. Bugs can be in one of several states: o - open A problem report has been submitted, no sanity checking performed. a - analyzed The problem is understood and a solution is being sought. f - feedback Further work requires additional information from the originator or the community - possibly confirmation of the effectiveness of a proposed solution. p - patched A patch has been committed, but some issues (MFC and / or confirmation from originator) are still open. r - repocopy The resolution of the problem report is dependent on a repocopy operation within the CVS repository which is awaiting completion. s - suspended The problem is not being worked on, due to lack of information or resources. This is a prime candidate for somebody who is looking for a project to do. If the problem cannot be solved at all, it will be closed, rather than suspended. c - closed A problem report is closed when any changes have been integrated, documented, and tested -- or when fixing the problem is abandoned. From bugmaster at FreeBSD.org Mon Sep 15 15:18:46 2008 From: bugmaster at FreeBSD.org (FreeBSD bugmaster) Date: Mon Sep 15 15:19:26 2008 Subject: Current problem reports assigned to freebsd-embedded@FreeBSD.org Message-ID: <200809151518.m8FFIjRQ018834@freefall.freebsd.org> Note: to view an individual PR, use: http://www.freebsd.org/cgi/query-pr.cgi?pr=(number). The following is a listing of current problems submitted by FreeBSD users. These represent problem reports covering all versions including experimental development code and obsolete releases. S Tracker Resp. Description -------------------------------------------------------------------------------- o kern/101228 embedded [nanobsd] [patch] Two more entries for FlashDevice.sub o misc/52256 embedded [picobsd] picobsd build script does not read in user/s o kern/42728 embedded [picobsd] many problems in src/usr.sbin/ppp/* after c o misc/15876 embedded [picobsd] PicoBSD message of the day problems 4 problems total. From smithi at nimnet.asn.au Tue Sep 16 17:26:18 2008 From: smithi at nimnet.asn.au (Ian Smith) Date: Tue Sep 16 17:26:25 2008 Subject: snooping an external 400kHz+ i2cbus Message-ID: <20080917011734.E439@sola.nimnet.asn.au> Hi, Maybe off topic here, being about using i386 laptop/s to both monitor and configure custom embedded devices that are programmed in assembler, but if anyone can help with this, I figure they're most likely here .. An engineer friend and I have designed a little system that has one AVR Tiny processor on either side of an electrically-isolated board, communicating over an optically-isolated i2c-(hardware)compatible bus. The bulk of the traffic is small packets of feedback data about once per ms. Eventually using some multiplexing of i2c addressing to encode some commands for compressed packets, but that's later; happy to run regular I2C protocol during development and tuning. What we need is a device that can a) snoop and log bus traffic, from one or the other side of the board, b) run in master mode for configuring and dumping the AVRs while debugging and c) run in slave mode for periodic data acquisition and monitoring, at 400Khz bursts at least. I dived into looking at iicbus system and drivers a couple of years ago, thinking there must be a way to use lpbb at 400kHz, but the only master and slave implementation found was then pcf for isabus at only 100kHz. I know nowhere near enough about FreeBSD drivers (or even C) to write one, though I may be able to hack on some existing code, with guidance. Are there perhaps now any pccard/cardbus or maybe USB adaptors that can talk 400kHz or better I2C, that could conceivably be used for this job? And more to the point, has anyone got it/them to work with FreeBSD? I'd be thankful for any pointers, or crumbs .. cheers, Ian From karin at visti.de Wed Sep 17 06:30:57 2008 From: karin at visti.de (Karin Willers) Date: Wed Sep 17 06:31:05 2008 Subject: snooping an external 400kHz+ i2cbus In-Reply-To: <20080917011734.E439@sola.nimnet.asn.au> References: <20080917011734.E439@sola.nimnet.asn.au> Message-ID: <1221631535.5729.6.camel@ratlos> Ian, you can have a look at the Beagle I2C/SPI/MDIO Protocol Analyzer at http://www.totalphase.com/products/beagle_usb12/ Although I never used that thingie myself, I would consider buying it when the need arises. Greetings, Karin ----- On Wed, 2008-09-17 at 03:06 +1000, Ian Smith wrote: > Hi, > > Maybe off topic here, being about using i386 laptop/s to both monitor > and configure custom embedded devices that are programmed in assembler, > but if anyone can help with this, I figure they're most likely here .. > > An engineer friend and I have designed a little system that has one AVR > Tiny processor on either side of an electrically-isolated board, > communicating over an optically-isolated i2c-(hardware)compatible bus. > > The bulk of the traffic is small packets of feedback data about once per > ms. Eventually using some multiplexing of i2c addressing to encode some > commands for compressed packets, but that's later; happy to run regular > I2C protocol during development and tuning. > > What we need is a device that can a) snoop and log bus traffic, from one > or the other side of the board, b) run in master mode for configuring > and dumping the AVRs while debugging and c) run in slave mode for > periodic data acquisition and monitoring, at 400Khz bursts at least. > > I dived into looking at iicbus system and drivers a couple of years ago, > thinking there must be a way to use lpbb at 400kHz, but the only master > and slave implementation found was then pcf for isabus at only 100kHz. I > know nowhere near enough about FreeBSD drivers (or even C) to write one, > though I may be able to hack on some existing code, with guidance. > > Are there perhaps now any pccard/cardbus or maybe USB adaptors that can > talk 400kHz or better I2C, that could conceivably be used for this job? > And more to the point, has anyone got it/them to work with FreeBSD? > > I'd be thankful for any pointers, or crumbs .. > > cheers, Ian > _______________________________________________ > freebsd-embedded@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-embedded > To unsubscribe, send any mail to "freebsd-embedded-unsubscribe@freebsd.org" > From smithi at nimnet.asn.au Wed Sep 17 17:20:22 2008 From: smithi at nimnet.asn.au (Ian Smith) Date: Wed Sep 17 17:20:34 2008 Subject: snooping an external 400kHz+ i2cbus In-Reply-To: <1221631535.5729.6.camel@ratlos> References: <20080917011734.E439@sola.nimnet.asn.au> <1221631535.5729.6.camel@ratlos> Message-ID: <20080918012553.P439@sola.nimnet.asn.au> On Wed, 17 Sep 2008, Karin Willers wrote: > you can have a look at the Beagle I2C/SPI/MDIO Protocol Analyzer at > > http://www.totalphase.com/products/beagle_usb12/ > > Although I never used that thingie myself, I would consider buying it > when the need arises. Thanks Karin, nice site and all. That or perhaps even their Aardvark http://www.totalphase.com/products/aardvark_i2cspi/ might do the job, if there's any chance of their linux software working, or being made to, and then if it works with our usb (gulp?). The xml scripting looks interesting at first glance. Aardvark only snoops up to 125kHz but looks good for master or slave, and may be the only one to work with USB 1.1 ports, ie both my laptops. Grabbed everything in sight, so have days of reading ahead. If the code and hardware is up to the quality of the docs, it looks like a winner at those prices. Lots of other good links from there too. cheers, Ian From gamuso at gmail.com Mon Sep 22 10:43:55 2008 From: gamuso at gmail.com (Odd-Jarle Kristoffersen) Date: Mon Sep 22 10:43:59 2008 Subject: Wi-Fi hotspot simple authentication Message-ID: <78AB86ED-28D8-4609-8A30-FB5060532DE5@gmail.com> Hello, Not sure which FreeBSD mailing-list this was most suited for, so I am trying here first. Been a FreeBSD-user for some years now, and mainly used it for firewall and web-hosting applications. I seem to recall stumbling across some information once upon a time about how to set up a simple WiFi access point that required a username and/or password web-based logon to access the Internet... What I want to accomplish is having a FreeBSD box wither with a built- in WiFi AP card, or with two ethernet interfaces where one is connected to a WiFi AP box. The other ethernet connected to a DSL- line. The WiFi should not have any WEP/WPA or any type of encryption. It should be completely open. Now for the tricky part. I want everyone who connects to this hotspot / AP to get a welcome webpage no matter which website they try to access at first. On this webpage they can just click a button to verify they read the disclaimer and gain full Internet access for the rest of their session. This system is intended for a non-profit guest harbor application that just wants to provide some basic information about the premises and the available local services, and provide any boat guests with free Internet during their stay. Is something like this at all possible using some spare computer parts and FreeBSD? Any exisiting boot-cd or similar systems out there already that you know of? Thanks Odd-Jarle From bugmaster at FreeBSD.org Mon Sep 22 11:06:51 2008 From: bugmaster at FreeBSD.org (FreeBSD bugmaster) Date: Mon Sep 22 11:07:03 2008 Subject: Current problem reports assigned to freebsd-embedded@FreeBSD.org Message-ID: <200809221106.m8MB6odB015322@freefall.freebsd.org> Note: to view an individual PR, use: http://www.freebsd.org/cgi/query-pr.cgi?pr=(number). The following is a listing of current problems submitted by FreeBSD users. These represent problem reports covering all versions including experimental development code and obsolete releases. S Tracker Resp. Description -------------------------------------------------------------------------------- o kern/101228 embedded [nanobsd] [patch] Two more entries for FlashDevice.sub o misc/52256 embedded [picobsd] picobsd build script does not read in user/s o kern/42728 embedded [picobsd] many problems in src/usr.sbin/ppp/* after c o misc/15876 embedded [picobsd] PicoBSD message of the day problems 4 problems total. From richard at unixguru.nl Mon Sep 22 12:23:20 2008 From: richard at unixguru.nl (Richard Arends) Date: Mon Sep 22 12:23:26 2008 Subject: Wi-Fi hotspot simple authentication In-Reply-To: <78AB86ED-28D8-4609-8A30-FB5060532DE5@gmail.com> References: <78AB86ED-28D8-4609-8A30-FB5060532DE5@gmail.com> Message-ID: <20080922111826.GB876@shell.unixguru.nl> On Mon, Sep 22, 2008 at 12:20:58PM +0200, Odd-Jarle Kristoffersen wrote: Hi, > Now for the tricky part. I want everyone who connects to this > hotspot / AP to get a welcome webpage no matter which website they try > to access at first. On this webpage they can just click a button to > verify they read the disclaimer and gain full Internet access for the > rest of their session. See pfsense (www.pfsense.com) -- Regards, Richard. /* Homo Sapiens non urinat in ventum */ From outbackdingo at gmail.com Mon Sep 22 13:48:27 2008 From: outbackdingo at gmail.com (Outback Dingo) Date: Mon Sep 22 13:48:29 2008 Subject: Wi-Fi hotspot simple authentication In-Reply-To: <20080922111826.GB876@shell.unixguru.nl> References: <78AB86ED-28D8-4609-8A30-FB5060532DE5@gmail.com> <20080922111826.GB876@shell.unixguru.nl> Message-ID: <5635aa0d0809220627r1dcfac4fw4897e96f4d346a8a@mail.gmail.com> yes PfSense will work, also see coova-chilli as ther eis a FreeBSd port for this also On Mon, Sep 22, 2008 at 6:18 PM, Richard Arends wrote: > On Mon, Sep 22, 2008 at 12:20:58PM +0200, Odd-Jarle Kristoffersen wrote: > > Hi, > > > Now for the tricky part. I want everyone who connects to this > > hotspot / AP to get a welcome webpage no matter which website they try > > to access at first. On this webpage they can just click a button to > > verify they read the disclaimer and gain full Internet access for the > > rest of their session. > > See pfsense (www.pfsense.com) > > -- > Regards, > > Richard. > > /* Homo Sapiens non urinat in ventum */ > _______________________________________________ > freebsd-embedded@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-embedded > To unsubscribe, send any mail to "freebsd-embedded-unsubscribe@freebsd.org > " > From mike at jellydonut.org Mon Sep 22 15:03:28 2008 From: mike at jellydonut.org (Michael Proto) Date: Mon Sep 22 15:03:35 2008 Subject: Wi-Fi hotspot simple authentication In-Reply-To: <78AB86ED-28D8-4609-8A30-FB5060532DE5@gmail.com> References: <78AB86ED-28D8-4609-8A30-FB5060532DE5@gmail.com> Message-ID: <1de79840809220730q740ff506l70790919d553192a@mail.gmail.com> On Mon, Sep 22, 2008 at 6:20 AM, Odd-Jarle Kristoffersen wrote: > Hello, > > Not sure which FreeBSD mailing-list this was most suited for, so I am > trying here first. > > Been a FreeBSD-user for some years now, and mainly used it for firewall and > web-hosting applications. > > I seem to recall stumbling across some information once upon a time about > how to set up a simple WiFi access point that required a username and/or > password web-based logon to access the Internet... > > What I want to accomplish is having a FreeBSD box wither with a built-in > WiFi AP card, or with two ethernet interfaces where one is connected to a > WiFi AP box. The other ethernet connected to a DSL-line. The WiFi should not > have any WEP/WPA or any type of encryption. It should be completely open. > > Now for the tricky part. I want everyone who connects to this hotspot / AP > to get a welcome webpage no matter which website they try to access at > first. On this webpage they can just click a button to verify they read the > disclaimer and gain full Internet access for the rest of their session. > > This system is intended for a non-profit guest harbor application that just > wants to provide some basic information about the premises and the available > local services, and provide any boat guests with free Internet during their > stay. > > Is something like this at all possible using some spare computer parts and > FreeBSD? Any exisiting boot-cd or similar systems out there already that you > know of? > > Thanks > Odd-Jarle > _______________________________________________ > freebsd-embedded@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-embedded > To unsubscribe, send any mail to "freebsd-embedded-unsubscribe@freebsd.org > " > Monowall will also do this, and has a captive portal interface available (pfSense was originally based on Monowall). http://m0n0.ch/wall/ -Proto From info at martenvijn.nl Mon Sep 22 17:33:54 2008 From: info at martenvijn.nl (Marten Vijn) Date: Mon Sep 22 17:33:57 2008 Subject: Wi-Fi hotspot simple authentication In-Reply-To: <1de79840809220730q740ff506l70790919d553192a@mail.gmail.com> References: <78AB86ED-28D8-4609-8A30-FB5060532DE5@gmail.com> <1de79840809220730q740ff506l70790919d553192a@mail.gmail.com> Message-ID: <1222104831.6016.2.camel@mvn-desktop> On Mon, 2008-09-22 at 10:30 -0400, Michael Proto wrote: > > > Monowall will also do this, and has a captive portal interface available > (pfSense was originally based on Monowall). > > http://m0n0.ch/wall/ > make sure you the beta version http://m0n0.ch/wall/beta.php stable is still running on FreeBSD 4.11 (unsported) and has crappy wifisupport Marten > > -Proto > _______________________________________________ > freebsd-embedded@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-embedded > To unsubscribe, send any mail to "freebsd-embedded-unsubscribe@freebsd.org" -- Marten Vijn http://martenvijn.nl http://opencommunitycamp.org http://wifisoft.org From gamuso at gmail.com Tue Sep 23 19:25:09 2008 From: gamuso at gmail.com (Odd-Jarle Kristoffersen) Date: Tue Sep 23 19:25:19 2008 Subject: Wi-Fi hotspot simple authentication In-Reply-To: <5635aa0d0809220627r1dcfac4fw4897e96f4d346a8a@mail.gmail.com> References: <78AB86ED-28D8-4609-8A30-FB5060532DE5@gmail.com> <20080922111826.GB876@shell.unixguru.nl> <5635aa0d0809220627r1dcfac4fw4897e96f4d346a8a@mail.gmail.com> Message-ID: <97DABA33-4AAD-4EE1-93D4-3024B099D1E4@gmail.com> Got plenty to test out from all these great options you all provided. I have started out with pfSense and it seems pretty nice. Will try out all the others as well. Thanks! Great input. Regards, Odd-Jarle Den 22. sep.. 2008 kl. 15.27 skrev Outback Dingo: > yes PfSense will work, also see coova-chilli as ther eis a FreeBSd > port for this also > > On Mon, Sep 22, 2008 at 6:18 PM, Richard Arends > wrote: > On Mon, Sep 22, 2008 at 12:20:58PM +0200, Odd-Jarle Kristoffersen > wrote: > > Hi, > > > Now for the tricky part. I want everyone who connects to this > > hotspot / AP to get a welcome webpage no matter which website they > try > > to access at first. On this webpage they can just click a button to > > verify they read the disclaimer and gain full Internet access for > the > > rest of their session. > > See pfsense (www.pfsense.com) > > -- > Regards, > > Richard. > > /* Homo Sapiens non urinat in ventum */ > _______________________________________________ > freebsd-embedded@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-embedded > To unsubscribe, send any mail to "freebsd-embedded-unsubscribe@freebsd.org > " > From info at martenvijn.nl Wed Sep 24 14:43:18 2008 From: info at martenvijn.nl (Marten Vijn) Date: Wed Sep 24 14:43:24 2008 Subject: Wi-Fi hotspot simple authentication In-Reply-To: <20080922111826.GB876@shell.unixguru.nl> References: <78AB86ED-28D8-4609-8A30-FB5060532DE5@gmail.com> <20080922111826.GB876@shell.unixguru.nl> Message-ID: <1222090032.23667.39.camel@polaris> On Mon, 2008-09-22 at 13:18 +0200, Richard Arends wrote: > On Mon, Sep 22, 2008 at 12:20:58PM +0200, Odd-Jarle Kristoffersen wrote: > > Hi, > > > Now for the tricky part. I want everyone who connects to this > > hotspot / AP to get a welcome webpage no matter which website they try > > to access at first. On this webpage they can just click a button to > > verify they read the disclaimer and gain full Internet access for the > > rest of their session. > > See pfsense (www.pfsense.com) Or if don't want a prefab solution and work (& learn) a bit more: for example: AP (soekris/alix) - nano/tinybsd - nocat/chillipot server - install on server radius apache2 on freebsd It's more work, but educating. kind regards Marten > -- Marten Vijn linux 2.0.18 OpenBSD 3.6 FreeBSD 4.6 http://martenvijn.nl http://opencommunitycamp.org http://wifisoft.org From bms at incunabulum.net Thu Sep 25 13:26:02 2008 From: bms at incunabulum.net (Bruce M Simpson) Date: Thu Sep 25 13:26:10 2008 Subject: ADM5120 for $34 Message-ID: <48DB8D94.5000308@incunabulum.net> JFYI: http://www.omnima.co.uk/store/catalog/Embedded-controller-p-16140.html They cite NetBSD support. 16MB DRAM is a bit small. 2MB NOR Flash is quite small. cheers BMS From bms at incunabulum.net Thu Sep 25 13:30:57 2008 From: bms at incunabulum.net (Bruce M Simpson) Date: Thu Sep 25 13:31:03 2008 Subject: Zyxel DMA-1000: Sigma (ARM7) based set top box Message-ID: <48DB928D.5000103@incunabulum.net> http://www.tomsguide.com/us/slideshow/Zyxel-DMA-1000-Media-Processor,0101-49318-2229----jpg-.html Believed to run ucLinux 2.4.x... retailing 75 UKP inc VAT here. From bob at sinister.com Sat Sep 27 00:52:34 2008 From: bob at sinister.com (Bob Keyes) Date: Sat Sep 27 00:52:40 2008 Subject: ADM5120 for $34 In-Reply-To: <48DB8D94.5000308@incunabulum.net> References: <48DB8D94.5000308@incunabulum.net> Message-ID: On Thu, 25 Sep 2008, Bruce M Simpson wrote: > JFYI: > http://www.omnima.co.uk/store/catalog/Embedded-controller-p-16140.html > > They cite NetBSD support. 16MB DRAM is a bit small. 2MB NOR Flash is quite > small. It is. Can NetBSD fit in this? It's difficult for Linux to do so, even with busybox and uclibc. Is there any more details about this board? Any US sources? Yes, I am looking for cheap netbsd platforms, with some provision for wifi cards. From imp at bsdimp.com Sat Sep 27 16:30:34 2008 From: imp at bsdimp.com (M. Warner Losh) Date: Sat Sep 27 16:30:41 2008 Subject: ADM5120 for $34 In-Reply-To: References: <48DB8D94.5000308@incunabulum.net> Message-ID: <20080927.103026.1723235596.imp@bsdimp.com> In message: Bob Keyes writes: : : : On Thu, 25 Sep 2008, Bruce M Simpson wrote: : : > JFYI: : > http://www.omnima.co.uk/store/catalog/Embedded-controller-p-16140.html : > : > They cite NetBSD support. 16MB DRAM is a bit small. 2MB NOR Flash is quite : > small. : : It is. Can NetBSD fit in this? It's difficult for Linux to do so, even : with busybox and uclibc. : : Is there any more details about this board? Any US sources? : : Yes, I am looking for cheap netbsd platforms, with some provision for wifi : cards. I bought this same box at CompUSA 2 years ago with case and power supply for $40.00... It is sitting on my desk right now waiting for some time to try the ADM5120 kernel on it... Warner From bms at incunabulum.net Sun Sep 28 22:14:55 2008 From: bms at incunabulum.net (Bruce M Simpson) Date: Sun Sep 28 22:15:02 2008 Subject: r181397 of src/usr.bin/make/job.c breaks cross-compiles Message-ID: <48E001DC.9010503@incunabulum.net> Hi Andrey, I see a change from you to job.c which is part of make... When cross-building a make as part of a NanoBSD arm job (amd64 7.1-PRERELEASE hosting an arm -CURRENT build), this change appears to break buildworld. If I back out this change and go back to rev 176808, the buildworld is fine. I don't understand what arc4random_uniform() is or its benefit -- I'm guessing it just acts as a uniform distribution filtering interface to arc4random() -- however the interface doesn't appear to present prior to -CURRENT, and this is breaking stuff which should work. Perhaps this change should be wrapped in an #if __FreeBSD_Version conditional so that it doesn't break 7.1->8.x source upgrades? thanks, BMS From ache at nagual.pp.ru Mon Sep 29 00:40:04 2008 From: ache at nagual.pp.ru (Andrey Chernov) Date: Mon Sep 29 00:40:11 2008 Subject: r181397 of src/usr.bin/make/job.c breaks cross-compiles In-Reply-To: <48E001DC.9010503@incunabulum.net> References: <48E001DC.9010503@incunabulum.net> Message-ID: <20080929002040.GA64397@nagual.pp.ru> On Sun, Sep 28, 2008 at 11:14:52PM +0100, Bruce M Simpson wrote: > When cross-building a make as part of a NanoBSD arm job (amd64 > 7.1-PRERELEASE hosting an arm -CURRENT build), this change appears to > break buildworld. If I back out this change and go back to rev 176808, > the buildworld is fine. Please explain, how it may cause problem? In 'make world' or 'make buildworld;make installworld' headers and libraries was builded first with old make, _then_ the new make is builded. 'make' is not a bootstrap tool builded first (at -current, I don't check -stable branch) If you uprade by hand you shound manually install headers, rebuild libs, then make. -- http://ache.pp.ru/ From bms at incunabulum.net Mon Sep 29 01:10:40 2008 From: bms at incunabulum.net (Bruce M Simpson) Date: Mon Sep 29 01:10:46 2008 Subject: r181397 of src/usr.bin/make/job.c breaks cross-compiles In-Reply-To: <20080929002040.GA64397@nagual.pp.ru> References: <48E001DC.9010503@incunabulum.net> <20080929002040.GA64397@nagual.pp.ru> Message-ID: <48E02B0D.9030806@incunabulum.net> Andrey Chernov wrote: > Please explain, how it may cause problem? In 'make world' or 'make > buildworld;make installworld' headers and libraries was builded first with > old make, _then_ the new make is builded. 'make' is not a bootstrap tool > builded first (at -current, I don't check -stable branch) > When I use NanoBSD to build a system image for the ARM, nanobsd.sh invokes the "make" targets in the source tree which I imagine most developers are familiar with; the build system attempts to build "make" first thing in the host's staging area under MAKEOBJDIRPREFIX, as one would normally expect. What is strange about this failure, is the fact that the "buildworld" target does not stop due to the error until after the dependency pass is made over the tree. I traced the problem back in logs to the "make" failing to find arc4random_uniform() during link. Given that it's building a "make" executable for the *host*, not the *target*, I would expect to see this error, as this function doesn't exist in FreeBSD 7.1. It looks to me as if the change to "make" assumes that the FreeBSD 8.x versions of the arc4 library functions are always available, even for cross compilation. This assumption doesn't seem to be robust, and it breaks down in this cross-compilation use case. Can you please advise what the best way to fix this in the source is? It looks as though the change should either be backed out, so the cross-compilation use case is unbroken, or wrapped with a __FreeBSD_version guard. thanks again for all your help BMS From ache at nagual.pp.ru Mon Sep 29 01:56:01 2008 From: ache at nagual.pp.ru (Andrey Chernov) Date: Mon Sep 29 01:56:07 2008 Subject: r181397 of src/usr.bin/make/job.c breaks cross-compiles In-Reply-To: <48E02B0D.9030806@incunabulum.net> References: <48E001DC.9010503@incunabulum.net> <20080929002040.GA64397@nagual.pp.ru> <48E02B0D.9030806@incunabulum.net> Message-ID: <20080929015554.GA66575@nagual.pp.ru> On Mon, Sep 29, 2008 at 02:10:37AM +0100, Bruce M Simpson wrote: > Andrey Chernov wrote: > > Please explain, how it may cause problem? In 'make world' or 'make > > buildworld;make installworld' headers and libraries was builded first with > > old make, _then_ the new make is builded. 'make' is not a bootstrap tool > > builded first (at -current, I don't check -stable branch) > > > > When I use NanoBSD to build a system image for the ARM, nanobsd.sh > invokes the "make" targets in the source tree which I imagine most > developers are familiar with; the build system attempts to build "make" > first thing in the host's staging area under MAKEOBJDIRPREFIX, as one > would normally expect. I can't find 'make' in -stable 7 and 6 boostrap tools too. The problem lies in the section which decide to build it first. It is absolutely not needed. Could you please show log or trace down to the section which builds 'make' first? > What is strange about this failure, is the fact that the "buildworld" > target does not stop due to the error until after the dependency pass is > made over the tree. Nothing strange, it works as I explain in the first reply.. > It looks to me as if the change to "make" assumes that the FreeBSD 8.x > versions of the arc4 library functions are always available, even for > cross compilation. This assumption doesn't seem to be robust, and it > breaks down in this cross-compilation use case. It not breaks because all is builded by old 'make' in 6, 7, and 8. New 'make' is not installed until the very final stage. I see no problem with cross-compilation in that way. > Can you please advise what the best way to fix this in the source is? I need to know more specifically the place which builds 'make' first to think about the fix. -- http://ache.pp.ru/ From ache at nagual.pp.ru Mon Sep 29 02:06:08 2008 From: ache at nagual.pp.ru (Andrey Chernov) Date: Mon Sep 29 02:06:14 2008 Subject: r181397 of src/usr.bin/make/job.c breaks cross-compiles In-Reply-To: <20080929015554.GA66575@nagual.pp.ru> References: <48E001DC.9010503@incunabulum.net> <20080929002040.GA64397@nagual.pp.ru> <48E02B0D.9030806@incunabulum.net> <20080929015554.GA66575@nagual.pp.ru> Message-ID: <20080929020606.GA67341@nagual.pp.ru> On Mon, Sep 29, 2008 at 05:55:56AM +0400, Andrey Chernov wrote: > > It looks to me as if the change to "make" assumes that the FreeBSD 8.x > > versions of the arc4 library functions are always available, even for > > cross compilation. This assumption doesn't seem to be robust, and it > > breaks down in this cross-compilation use case. > > It not breaks because all is builded by old 'make' in 6, 7, and 8. New > 'make' is not installed until the very final stage. I see no problem with > cross-compilation in that way. I just check cross-tools list in 6, 7 and 8 and not found 'make' there too. -- http://ache.pp.ru/ From bms at incunabulum.net Mon Sep 29 02:41:52 2008 From: bms at incunabulum.net (Bruce M Simpson) Date: Mon Sep 29 02:41:59 2008 Subject: r181397 of src/usr.bin/make/job.c breaks cross-compiles In-Reply-To: <20080929015554.GA66575@nagual.pp.ru> References: <48E001DC.9010503@incunabulum.net> <20080929002040.GA64397@nagual.pp.ru> <48E02B0D.9030806@incunabulum.net> <20080929015554.GA66575@nagual.pp.ru> Message-ID: <48E0406E.3030409@incunabulum.net> Andrey Chernov wrote: > I need to know more specifically the place which builds 'make' first > to think about the fix. > The failure occurs during the "Building an up-to-date make" stage of buildworld, it is being invoked by nanobsd.sh. From ru at FreeBSD.org Mon Sep 29 06:30:34 2008 From: ru at FreeBSD.org (Ruslan Ermilov) Date: Mon Sep 29 06:30:40 2008 Subject: r181397 of src/usr.bin/make/job.c breaks cross-compiles In-Reply-To: <20080929020606.GA67341@nagual.pp.ru> References: <48E001DC.9010503@incunabulum.net> <20080929002040.GA64397@nagual.pp.ru> <48E02B0D.9030806@incunabulum.net> <20080929015554.GA66575@nagual.pp.ru> <20080929020606.GA67341@nagual.pp.ru> Message-ID: <20080929061057.GB57821@edoofus.dev.vega.ru> Hi, On Mon, Sep 29, 2008 at 06:06:06AM +0400, Andrey Chernov wrote: > On Mon, Sep 29, 2008 at 05:55:56AM +0400, Andrey Chernov wrote: > > > It looks to me as if the change to "make" assumes that the FreeBSD 8.x > > > versions of the arc4 library functions are always available, even for > > > cross compilation. This assumption doesn't seem to be robust, and it > > > breaks down in this cross-compilation use case. > > > > It not breaks because all is builded by old 'make' in 6, 7, and 8. New > > 'make' is not installed until the very final stage. I see no problem with > > cross-compilation in that way. > > I just check cross-tools list in 6, 7 and 8 and not found 'make' there > too. > The first thing buildworld does is to check if a passed instance of `make' is good enough; if not, it's bootstrapped -- built and used later in the process. It's handled by the `upgrade_checks' target in src/Makefile which buildworld depends on. An ability to compile the latest make on the lowest supported version of FreeBSD is a requirement of our build procedure. If your change doesn't meet it, you should really think of either backing it out, or making it compile and work. For RELENG_7 it's currently FreeBSD 5.3, for HEAD -- FreeBSD 6.0. Cheers, -- Ruslan Ermilov ru@FreeBSD.org FreeBSD committer From bugmaster at FreeBSD.org Mon Sep 29 11:06:49 2008 From: bugmaster at FreeBSD.org (FreeBSD bugmaster) Date: Mon Sep 29 11:07:18 2008 Subject: Current problem reports assigned to freebsd-embedded@FreeBSD.org Message-ID: <200809291106.m8TB6mpH040751@freefall.freebsd.org> Note: to view an individual PR, use: http://www.freebsd.org/cgi/query-pr.cgi?pr=(number). The following is a listing of current problems submitted by FreeBSD users. These represent problem reports covering all versions including experimental development code and obsolete releases. S Tracker Resp. Description -------------------------------------------------------------------------------- o kern/101228 embedded [nanobsd] [patch] Two more entries for FlashDevice.sub o misc/52256 embedded [picobsd] picobsd build script does not read in user/s o kern/42728 embedded [picobsd] many problems in src/usr.sbin/ppp/* after c o misc/15876 embedded [picobsd] PicoBSD message of the day problems 4 problems total. From bms at incunabulum.net Mon Sep 29 11:32:01 2008 From: bms at incunabulum.net (Bruce M Simpson) Date: Mon Sep 29 11:32:07 2008 Subject: r181397 of src/usr.bin/make/job.c breaks cross-compiles In-Reply-To: <48E0406E.3030409@incunabulum.net> References: <48E001DC.9010503@incunabulum.net> <20080929002040.GA64397@nagual.pp.ru> <48E02B0D.9030806@incunabulum.net> <20080929015554.GA66575@nagual.pp.ru> <48E0406E.3030409@incunabulum.net> Message-ID: <48E0BCAE.9070508@incunabulum.net> Bruce M Simpson wrote: > Andrey Chernov wrote: >> I need to know more specifically the place which builds 'make' first >> to think about the fix. >> > > The failure occurs during the "Building an up-to-date make" stage of > buildworld, it is being invoked by nanobsd.sh. > I can't supply logs from the nanobsd build unfortunately, the error has disappeared since I manually backed out the rev from my SVN checkout. If I have time to spare, which is unlikely, I can try to reproduce w/logs. However, the error message is almost identical to an earlier report involving the tinderbox, so this may be enough for you to go on: http://kerneltrap.org/mailarchive/freebsd-current/2008/8/7/2855624 I believe Tinderbox in this case is also being used to drive a cross-compile. Ruslan has explained why and how "make" is being rebuilt, for 7.1->8.x cross compiles this is indeed what's happening. it would be great to get this sorted, it seems likely to bite other folk. cheers BMS From ache at nagual.pp.ru Mon Sep 29 15:43:27 2008 From: ache at nagual.pp.ru (Andrey Chernov) Date: Mon Sep 29 15:43:34 2008 Subject: r181397 of src/usr.bin/make/job.c breaks cross-compiles In-Reply-To: <48E0BCAE.9070508@incunabulum.net> References: <48E001DC.9010503@incunabulum.net> <20080929002040.GA64397@nagual.pp.ru> <48E02B0D.9030806@incunabulum.net> <20080929015554.GA66575@nagual.pp.ru> <48E0406E.3030409@incunabulum.net> <48E0BCAE.9070508@incunabulum.net> Message-ID: <20080929154320.GA87747@nagual.pp.ru> On Mon, Sep 29, 2008 at 12:31:58PM +0100, Bruce M Simpson wrote: > However, the error message is almost identical to an earlier report > involving the tinderbox, so this may be enough for you to go on: > > http://kerneltrap.org/mailarchive/freebsd-current/2008/8/7/2855624 > > I believe Tinderbox in this case is also being used to drive a > cross-compile. Ruslan has explained why and how "make" is being rebuilt, > for 7.1->8.x cross compiles this is indeed what's happening. This is completely different case and already fixed. strfile is in bootstrap tools but make is not. -- http://ache.pp.ru/ From ache at nagual.pp.ru Mon Sep 29 15:55:37 2008 From: ache at nagual.pp.ru (Andrey Chernov) Date: Mon Sep 29 15:55:50 2008 Subject: r181397 of src/usr.bin/make/job.c breaks cross-compiles In-Reply-To: <20080929061057.GB57821@edoofus.dev.vega.ru> References: <48E001DC.9010503@incunabulum.net> <20080929002040.GA64397@nagual.pp.ru> <48E02B0D.9030806@incunabulum.net> <20080929015554.GA66575@nagual.pp.ru> <20080929020606.GA67341@nagual.pp.ru> <20080929061057.GB57821@edoofus.dev.vega.ru> Message-ID: <20080929155533.GB87747@nagual.pp.ru> On Mon, Sep 29, 2008 at 10:10:57AM +0400, Ruslan Ermilov wrote: > The first thing buildworld does is to check if a passed instance of > `make' is good enough; if not, it's bootstrapped -- built and used > later in the process. It's handled by the `upgrade_checks' target > in src/Makefile which buildworld depends on. Ah, 'upgrade_checks' is the problem. Then simple ifdefing with __FreeBSD_version will be enough. -- http://ache.pp.ru/ From ache at nagual.pp.ru Mon Sep 29 16:14:33 2008 From: ache at nagual.pp.ru (Andrey Chernov) Date: Mon Sep 29 16:14:39 2008 Subject: r181397 of src/usr.bin/make/job.c breaks cross-compiles In-Reply-To: <48E0406E.3030409@incunabulum.net> References: <48E001DC.9010503@incunabulum.net> <20080929002040.GA64397@nagual.pp.ru> <48E02B0D.9030806@incunabulum.net> <20080929015554.GA66575@nagual.pp.ru> <48E0406E.3030409@incunabulum.net> Message-ID: <20080929161430.GC87747@nagual.pp.ru> On Mon, Sep 29, 2008 at 03:41:50AM +0100, Bruce M Simpson wrote: > Andrey Chernov wrote: > > I need to know more specifically the place which builds 'make' first > > to think about the fix. > > > > The failure occurs during the "Building an up-to-date make" stage of > buildworld, it is being invoked by nanobsd.sh. I just commit the fix. -- http://ache.pp.ru/ From bms at incunabulum.net Mon Sep 29 22:11:37 2008 From: bms at incunabulum.net (Bruce M Simpson) Date: Mon Sep 29 22:11:43 2008 Subject: r181397 of src/usr.bin/make/job.c breaks cross-compiles In-Reply-To: <20080929161430.GC87747@nagual.pp.ru> References: <48E001DC.9010503@incunabulum.net> <20080929002040.GA64397@nagual.pp.ru> <48E02B0D.9030806@incunabulum.net> <20080929015554.GA66575@nagual.pp.ru> <48E0406E.3030409@incunabulum.net> <20080929161430.GC87747@nagual.pp.ru> Message-ID: <48E15296.6090504@incunabulum.net> Andrey Chernov wrote: > On Mon, Sep 29, 2008 at 03:41:50AM +0100, Bruce M Simpson wrote: > >> The failure occurs during the "Building an up-to-date make" stage of >> buildworld, it is being invoked by nanobsd.sh. >> > > I just commit the fix. > > Thanks for taking the time to look at this, I really appreciate it. cheers, BMS