Re: [Firebird-devel] DatabaseAccess path prepended to remote connection strings
On Tue, Jul 24, 2012 at 10:19:20AM +0400, Alex Peshkoff wrote:
> Patch seems to be absolutely correct.
Thank you for the review. Committed to HEAD.
> Just interested - did it really try to open something like
> '/srv/firebird/172.16.114.128:tips'?
Yes, I found this in strace log:
("/usr/bin/isql", ["isql", "-ch", "UTF8", "172.16.114.1:tips"], [/* 55 vars
*/]) = 0
22767 access("/srv/firebird/172.16.114.1:tips", R_OK) = -1 EACCES (Permission
denied)
22767 sendto(4, "\2\0\0\0\20\0\0\0\36\0\0\0srv/firebird/172.16."..., 42,
MSG_NOSIGNAL, NULL, 0) = 42
22767 access("/srv/firebird/172.16.114.1:tips", R_OK) = -1 EACCES (Permission
denied)
22767 stat("/srv/firebird/172.16.114.1:tips", 0x76504e40) = -1 EACCES
(Permission denied)
22767 open("/srv/firebird/172.16.114.1:tips", O_RDWR) = -1 EACCES (Permission
denied)
22767 open("/srv/firebird/172.16.114.1:tips", O_RDONLY) = -1 EACCES (Permission
denied)
The sendto() call is a query sent to nscd, probably connected to error
message "Undefined service srv/firebird/172.16.114.1/tcp.". Apparently
the client also parsed the string "/srv/firebird/172.16.114.1:tips"
according to the "host/port:database" scheme.
Michal Kubeček
--
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
Firebird-Devel mailing list, web interface at
https://lists.sourceforge.net/lists/listinfo/firebird-devel
[Firebird-devel] [FB-Tracker] Created: (CORE-3897) Access violation. The code attempted to access a virtual address without privilege to do so. This exception will cause the Firebird server to termina
Access violation. The code attempted to access a virtual address without privilege to do so. This exception will cause the Firebird server to terminate abnormally. -- Key: CORE-3897 URL: http://tracker.firebirdsql.org/browse/CORE-3897 Project: Firebird Core Issue Type: Bug Affects Versions: 2.5.1 Environment: Windows Sever 2003, Enterprise Edition, Service Pack 2 Reporter: maria oprea Firebird 2.5.1 superclassic (win32) crash. I restart firebird but it crash again. I install firebird again and it didn't crash. In firebird.log I have: Access violation. The code attempted to access a virtual address without privilege to do so. This exception will cause the Firebird server to terminate abnormally. INET/inet_error: read errno = 10093 INET/inet_error: select in packet_receive errno = 10093 INET/inet_error: read errno = 10054 INET/select_wait: select failed, errno = 10093 SRVR_multi_thread/RECEIVE: error on main_port, shutting down -- This message is automatically generated by JIRA. - If you think it was sent incorrectly contact one of the administrators: http://tracker.firebirdsql.org/secure/Administrators.jspa - For more information on JIRA, see: http://www.atlassian.com/software/jira -- Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
Re: [Firebird-devel] Building Firebird for Android
On 20/07/2012 15:09, Alex Peshkoff wrote: > On 07/20/12 16:20, JLM wrote: > Checkout trunk and use the following configure (autogen.sh) parameter > --with-cross-build=android.arme > > BTW - I did not find any VM in android's NDK. What is the best way to > check/debug android's soft? > Please also check correctness of defines in > src/include/cross/android.arme - for lear reasons I had to emulate > autogen's output manually. > -- > Live Security Virtual Conference > Exclusive live event will cover all the ways today's security and > threat landscape has changed and how IT managers can respond. Discussions > will include endpoint security, mobile security and the latest in malware > threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ > Firebird-Devel mailing list, web interface at > https://lists.sourceforge.net/lists/listinfo/firebird-devel Hello, I don't understand : yesterday I succeed in compiling libfbclient and today after I had checkout the trunk again, it used to run. It says "configure: error: cannot run C compiled programs. If you meant to cross compile, use `--host'." And configure seems to not read the android.arme file, I can try any name of file, it doesn't look in the file. Same thing when I put --host=arm-linux-androideabi in parameters of the configure, it stop at "checking whether setpgrp takes no argument... configure: error: cannot check setpgrp when cross compiling" whereas the android.arme file is there to fix this kind of problem. Regards, Jean-Louis Meuriot -- Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
Re: [Firebird-devel] Building Firebird for Android
On 07/24/12 16:56, JLM wrote: > On 20/07/2012 15:09, Alex Peshkoff wrote: >> On 07/20/12 16:20, JLM wrote: >> Checkout trunk and use the following configure (autogen.sh) parameter >> --with-cross-build=android.arme >> >> BTW - I did not find any VM in android's NDK. What is the best way to >> check/debug android's soft? >> Please also check correctness of defines in >> src/include/cross/android.arme - for lear reasons I had to emulate >> autogen's output manually. >> -- >> Live Security Virtual Conference >> Exclusive live event will cover all the ways today's security and >> threat landscape has changed and how IT managers can respond. Discussions >> will include endpoint security, mobile security and the latest in malware >> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ >> Firebird-Devel mailing list, web interface at >> https://lists.sourceforge.net/lists/listinfo/firebird-devel > Hello, > > I don't understand : yesterday I succeed in compiling libfbclient and > today after I had checkout the trunk again, it used to run. > It says "configure: error: cannot run C compiled programs. > If you meant to cross compile, use `--host'." > > And configure seems to not read the android.arme file, I can try any > name of file, it doesn't look in the file. > Same thing when I put --host=arm-linux-androideabi in parameters of the > configure, it stop at > "checking whether setpgrp takes no argument... configure: error: cannot > check setpgrp when cross compiling" > whereas the android.arme file is there to fix this kind of problem. Sooner of all you've used slightly different switches in configure. BTW, --host should not be used. -- Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
Re: [Firebird-devel] Building Firebird for Android
On 07/23/12 17:48, Adriano dos Santos Fernandes wrote: > On 23/07/2012 10:34, Alex Peshkoff wrote: >> On 07/23/12 14:41, Adriano dos Santos Fernandes wrote: >>> On 23/07/2012 06:49, Alex Peshkoff wrote: Somewhy it happened that in my trunk I have platform.h in extern/icu/source/common/unicode. No idea how did it arrive in that dir (certainly, not under version control). As far as I know about Android, icu libraries are present in default device's setup. But question comes - what is best way to access them when doing cross-build? >>> If it's not in the NDK, then the library is unsupported and may not be >>> in all versions. It's there just to the OS. >>> >>> So in this case we should supply it. But for client only we would not >>> need ICU. This is a problem of our "common" library who have server-code >>> which is never called by the client. >> Hmm... >> I do remember that I had to add CharSet.cpp to common in order to build >> client. Will recheck. >> >> > The problem is that some things (specially unicode_util) was added to > common, but in fact, it's a piece needed only in engine and intl libraries. > Do you want to say that a member of static library, not used to resolve symbols during link, will anyway cause errors when it references non-existent library? I've always thought that ld is not so foolish. -- Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
Re: [Firebird-devel] Building Firebird for Android
On 24/07/2012 10:15, Alex Peshkoff wrote: >> The problem is that some things (specially unicode_util) was added to >> common, but in fact, it's a piece needed only in engine and intl libraries. >> > Do you want to say that a member of static library, not used to resolve > symbols during link, will anyway cause errors when it references > non-existent library? I've always thought that ld is not so foolish. > > No, as we saw, it is working in trunk. It's not the greatest design, but... As I pointed it out, Android SDK/NDK has no ICU library and then our client (which technically does not need ICU) does not build without first manually build/install ICU. Weird. Adriano -- Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
Re: [Firebird-devel] Conversion of CVS to Subversion
As there are not replies as of yet, can I assume there is no problem to do this migration on Sunday 2012-08-05 (August fifth) between 12:00 and 16:00 UTC (14:00 and 18:00 Central European Summer Time)? Mark On 22-7-2012 15:25, Mark Rotteveel wrote: > I finished the instructions and tested them (except for the > sourceforge-specific parts) on a local copy of the Firebird-repository. > See below. For now I have assumed to use my own username for cvs2svn > specific revisions. > > During the import the repository is not available for commits. The > creation of the dumpfile is can be done in advance, assuming no further > checkins happen to the client-java module in CVS. > > I guess that the migration will take 1 to 2 hours, assuming I use rsync > to update the backup of the svn repository on my machine instead of a > full new backup (1-2 minutes to update between yesterday and today, > about 1.5 hours for a new, full rsync). > There however is uncertainty on the amount of time that copying the > repository for edit (step 3b) and saving it back (step 5a) will take and > how fast loading the dumpfile into the repository is on sourceforge > (step 4; this takes about 5 minutes on my machine). > > The timetable for this would be: > T - 1 hour : reminder to firebird-devel about migration > T - 5 minutes : announcement to firebird-devel that migration starts, > no more subversion commits from this time. > T + 0 : Backup repository (step 1) > T + 5 minutes : Create client-java folder (step 2) > T + 6 minutes : Upload jaybirddump.zip to sourceforge and unzip (step 3a) > T + 15 minutes : Checkout repository (step 3b) > T + 45 minutes : Load dumpfile into repository (step 4) > T + 75 minutes : Save repository back (step 5a) > T + 105 minutes: Migration done, sanity checks on repository (eg browse, > checkout) > T + 110 minutes: announcement to firebird-devel, subversion commits > allowed again > > Note: I have assumed 30 minutes for each step with uncertainty in time, > if things go faster, then the repository will simply be available > sooner. However we probably should factor in an additional two hours > just to be on the safe side. > > I propose to do this migration on Sunday 2012-08-05 (August fifth) > between 12:00 and 16:00 UTC (14:00 and 18:00 Central European Summer > Time), as it seems that Sundays are usually pretty low on commits. > > Rollback before step 5a will consist of simply calling adminrepo --discard. > > Rollback after step 5a consists of me generating a dumpfile based on the > backup and recreating and reloading the repository using the > instructions on > http://sourceforge.net/apps/trac/sourceforge/wiki/SVN%20adminrepo#Importingthedumpfile > Given the size of the repository that rollback scenario would probably > take several hours. > > Steps to import Jaybird into Subversion > === > > Instructions based on Windows and Cygwin with folder /repomigration for > various backups and files. > Migration uses trunk version of cvs2svn from cvs2svn.tigris.org (and not > the cygwin version) to solve problems with linebreaks(?) in log texts. > > Create the dumpfile from CVS > > > To prevent conflict with local (Windows) python: > unset PYTHONHOME > > 1) backup the entire CVS repository using rsync: (execute from folder > /repomigration/cvsrepo) > rsync -av firebird.cvs.sourceforge.net::cvsroot/firebird/* . > 2) Create dumpfile > /repomigration/cvs2svn-trunk/cvs2svn --username=mrotteveel > --auto-props=/repomigration/svn-eol-style.txt > --mime-types=/repomigration/mime.types -–eol-from-mime-type > --dumpfile=/repomigration/jaybirddump /repomigration/cvsrepo/client-java > > svn-eol-style.txt:http://www.apache.org/dev/svn-eol-style.txt > mime.types: > http://svn.apache.org/repos/asf/httpd/httpd/trunk/docs/conf/mime.types > > Import the dumpfile into SVN > > > Note: during import, no commits can be made by other users > > 1) Backup the SVN repository using rsync: (execute from folder > /repomigration/svnrepo) > rsync -av firebird.svn.sourceforge.net::svn/firebird/* . > 2) Create client-java folder in the subversion repository (eg using > TortoiseSVN) or commandline: > svn mkdir -m "Create repository-folder for Jaybird (Firebird JDBC)" > https://firebird.svn.sourceforge.net/svnroot/firebird/client-java > 3) 3a) Upload zipped dumpfile and unzip: > Use sf-help to find out the shell session and host info for sftp to > connect to > Use sftp (or WinSCP) to upload the dumpfile (to > /home/project-web/firebird/htdocs as that seems to be the only writable > location according to most sources) > 3b) Checkout repository for edit into /svnroot/firebird: > adminrepo --checkout svn > 4) Load dump into repository into folder client-java: > svnadmin load --parent-dir client-java /svnroot/firebird < jaybirddump > 5) 5a) Save checkedout repository back: >
Re: [Firebird-devel] Conversion of CVS to Subversion
On 07/24/12 22:25, Mark Rotteveel wrote: > As there are not replies as of yet, can I assume there is no problem to > do this migration on Sunday 2012-08-05 (August fifth) > between 12:00 and 16:00 UTC (14:00 and 18:00 Central European Summer > Time)? OK for me. -- Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
