October 2006 - developers-list

[Monetdb-developers] [Monetdb-sql-checkins] Problems with SQL installation.
by Stefan Manegold 10 Oct '06

10 Oct '06

Romulo, this should have been either filed as a bug report or sent to Monetdb-developers(a)lists.sourceforge.net. On Tue, Oct 10, 2006 at 05:19:32PM +0200, Romulo Goncalves wrote: > Dear SQL developers, > > I checked out the buildtools and I installed it. > > > I checked out MonetDB4 v4.13.1 and I install with this configure: > configure --prefix=/ufs/goncalve/scratch/MonetDB/monetdb4 --enable-debug > --enable-bits=64 At this point: is /ufs/goncalve/scratch/MonetDB/monetdb4/bin in your PATH? > I checked out MonetDB5 version v4.99.19 and I installed it with the > following configure: > configure --prefix=/ufs/goncalve/scratch/MonetDB/monetdb5 --enable-debug > --enable-bits=64 > > > Then I checked out SQL module and I installed it with the following > configure: At this point: is /ufs/goncalve/scratch/MonetDB/monetdb5/bin in your PATH? > configure --prefix=/ufs/goncalve/scratch/MonetDB/monetdb5/ > --enable-debug --enable-bits=64 > > > For all the installations the directories were removed and the > bootstraap was executed. > > The error that I got in the installation of SQL is: > > ve/scratch/MonetDB/monetdb4/lib/libstream.so -lpthread -m64 -Wl,-soname > -Wl,lib _sql.so.0 -o .libs/lib_sql.so.0.0.0 > /usr/bin/ld: cannot find -loptimizer > collect2: ld returned 1 exit status > make[7]: *** [lib_sql.la] Error 1 > make[7]: Leaving directory > `/net/amelia/export/scratch1/goncalve/SQL/sql/build/s rc/backends/monet5' > > > This problem started after the following update: > > [Monetdb-sql-checkins] sql/src/backends/monet5 Makefile.ag, 1.30, 1.31 > > Update of /cvsroot/monetdb/sql/src/backends/monet5 > In directory > sc8-pr-cvs7.sourceforge.net:/tmp/cvs-serv6378/src/backends/monet5 > > Modified Files: > Makefile.ag > Log Message: > Added some libraries that are used by the module. did you re-compiler Monet5 & SQL from scratch (bootstrap; configure; make; make install) after this update? > Could you tell me where I am wrong? > > Best regards, > Romulo Goncalves Stefan -- | Dr. Stefan Manegold | mailto:Stefan.Manegold@cwi.nl | | CWI, P.O.Box 94079 | http://www.cwi.nl/~manegold/ | | 1090 GB Amsterdam | Tel.: +31 (20) 592-4212 | | The Netherlands | Fax : +31 (20) 592-4312 |

3 4

[Monetdb-developers] Linking against a library
by Wouter Scherphof 09 Oct '06

09 Oct '06

Hi, My first module (called "geom"), created from "template", compiles (on Windows XP with the "native tools") and the hello() function works. Now I want to use some functions of a certain library (the GEOS<http://geos.refractions.net>C API). I'm fairly new to C, so probably I'm doing something obvious wrong. What I have is a header file "geos_c.h<http://geos.refractions.net/ro/doxygen_docs/html/geos__c_8h-source.html>" and a DLL "geos_c_fw.dll" (I chose to use a prebuilt DLL for now, since I didn't manage to build the GEOS library as well :-(.) I put those in the NT directory. In my geom.mx, I added: *#include* <geos_c.h>* **int* testGeos( ){* * initGEOS(*NULL*,*NULL*);* * finishGEOS();* ** return* GDK_SUCCEED;* *} Now, when I run "nmake NEED_MX=1", I get: (...) cl -GF -W3 -wd4273 -wd4102 -MD -nologo -Zi -Ox -I. -I.\.. "-IC:\Pthreads\include" -I.\..\..\src "-IC:\monetdb\MonetDB\NT\include" "-IC:\monetd b\MonetDB\NT\include\common" "-IC:\monetdb\MonetDB\NT\include\gdk" "-IC:\monetdb\MonetDB\NT\include\monet" "-IC:\monetdb\MonetDB\NT\include\plain" "-I C:\monetdb\MonetDB\NT\include\mapi" -DNDEBUG -DLIBGEOM -DCOMPILE_DL_GEOM -Fo"geom.obj" -c "geom.c" geom.c cl -GF -W3 -wd4273 -wd4102 -MD -nologo -Zi -Ox -I. -I.\.. "-IC:\Pthreads\include" -I.\..\..\src "-IC:\monetdb\MonetDB\NT\include" "-IC:\monetd b\MonetDB\NT\include\common" "-IC:\monetdb\MonetDB\NT\include\gdk" "-IC:\monetdb\MonetDB\NT\include\monet" "-IC:\monetdb\MonetDB\NT\include\plain" "-I C:\monetdb\MonetDB\NT\include\mapi" -DNDEBUG -LD -Felib_geom.dll geom.glue.obj geom.obj /link "/LIBPATH:C:\monetdb\MonetDB\NT\lib" "/LIBPATH:C:\monet db\MonetDB\NT\lib\MonetDB" libbat.lib libmonet.lib Creating library lib_geom.lib and object lib_geom.exp geom.obj : error LNK2019: unresolved external symbol _finishGEOS referenced in function _testGeos geom.obj : error LNK2019: unresolved external symbol _initGEOS referenced in function _testGeos lib_geom.dll : fatal error LNK1120: 2 unresolved externals So the compilation of geom.c succeeds, but the linking of lib_geom.dll fails. Apperently, the geos_c_fw.dll is not found. I tried putting it in one of the /LIBPATH, but that doesn't help. How can I make this work? Best regards, Wouter Scherphof. Utrecht Professional School, The Netherlands.

2 1

[Monetdb-developers] Crash leaves me unable to start mserver
by Joseph.Baird＠cox.com 09 Oct '06

09 Oct '06

After loading several million records via odbc, the server is suddenly unresponsive but can be restarted. When using jdbc, the server crashes, again after several million records, but cannot be restarted. Performance and usability are impressive with a million records or so but I cannot seem to load more than that without consistent problems. Is there any way to make the server - and ideally the database - usable again? Thank you in advance, Joe Baird

2 1

Re: [Monetdb-developers] Monetdb-developers Digest, Vol 5, Issue 6
by p.a.boncz 06 Oct '06

06 Oct '06

Hi all, > In PROC rpc_client() (runtime/xrpc.mx), I had changed > ws_addcoll(ws, ... > to > ws_opencoll(ws_id(ws), ... it should we ws_opencoll(wsid, ...) You have that, because in the query context we always have both variable 'ws' and 'wsid'. ws_id(ws) should be called once, it asks for a unique oid. The combination of (id,int(ws)) is squeezed into a lng, and that is wsid. As for doc_tbl(), what would work is to change its parameter to wsid; one can always get the ws with ws := ws_bat(wsid), this is done now in numerous places in pf_support.mx. Your conclusions (1), (2) and (3) are all correct. But, Jan's suggestion to store the wsid inside the ws can also work. However, I deem it very awkward to introduce a new BAT in the ws just to retain a number. Other suggestions are to rename the ws-bat to some unique key, and then we can use that name as transaction key (=wsid). We would change the wsid type then from lng to str. Another idea is to set the head-seqbase of the ws-bat to an oid. I think the ws is accessed with fetch() only, which disregards any seqbase. Note that internally in pathfinder.mx there are a number of functions that just pass around an artificial wsid, and in fact a ws does not exist at all (this is the MIL document management interface e.g. shred_doc), so a number of (internal) functions in pathfinder.mx will have to stay using a wsid in their signature. They do not need a ws-bat. But the exported functions ws_opencoll and ws_opendoc could indeed just use a ws as they did before. I will not have time for this until next week, so if anyone feels like adventurous, he may try. Peter -- | Dr. Stefan Manegold | mailto:Stefan.Manegold@cwi.nl | | CWI, P.O.Box 94079 | http://www.cwi.nl/~manegold/ | | 1090 GB Amsterdam | Tel.: +31 (20) 592-4212 | | The Netherlands | Fax : +31 (20) 592-4312 | ------------------------------ Message: 2 Date: Fri, 6 Oct 2006 12:43:21 +0200 From: Stefan Manegold <Stefan.Manegold(a)cwi.nl> Subject: Re: [Monetdb-developers] The new "ws-API", Algebra & XRpc (& PFtijah) To: Monetdb-developers(a)lists.sourceforge.net Message-ID: <20061006104321.GA2821(a)corona.ins.cwi.nl> Content-Type: text/plain; charset=us-ascii On Thu, Oct 05, 2006 at 11:03:09AM +0200, Stefan Manegold wrote: > Dear Peter, dear fellow PF & MXQ developers, [...] > In cases of XRpc & Algebra, there are still places were wsid := ws_id(ws) is > IMHO called in the wrong location: > > XRpc: > In PROC rpc_client() (runtime/xrpc.mx), I had changed > ws_addcoll(ws, ... > to > ws_opencoll(ws_id(ws), ... > While this seems to worg for now with read-only documents (i.e., in the > absense of updates and concurrency), it is IMHO incorrect in general. > Rather, rpc_client should receive wsid as argument instead of ws; > if necessary, ws can then be derived via ws := ws_bat(wsid). I just changed the signatures of doLoopLiftedRPC() & rpc_client() to receive lng wsid instead of BAT ws, BAT ws is the locally derived via ws := ws_bat(wsid); All XRpc tests seem to work fine. Stefan [...] -- | Dr. Stefan Manegold | mailto:Stefan.Manegold@cwi.nl | | CWI, P.O.Box 94079 | http://www.cwi.nl/~manegold/ | | 1090 GB Amsterdam | Tel.: +31 (20) 592-4212 | | The Netherlands | Fax : +31 (20) 592-4312 | ------------------------------ Message: 3 Date: Fri, 06 Oct 2006 14:39:19 +0200 From: Jan Rittinger <rittinge(a)in.tum.de> Subject: Re: [Monetdb-developers] The new "ws-API", Algebra & XRpc (& PFtijah) To: Monetdb-developers(a)lists.sourceforge.net Message-ID: <45264E77.3060201(a)in.tum.de> Content-Type: text/plain; charset=ISO-8859-1; format=flowed If this wsid is a concept that is strictly necessary for the update support (and nobody except me thinks that this bit shifting might be not the nicest solution) we could also think about getting rid of the ws and replacing it with the wsid. I just do not like the fact that there are two variables underway that have some side effects on each other. So a fourth proposal (d) would be to replace the occurrences of ws by wsid. This would mean that *all* document accesses would have to include another indirection (even all path steps). On 10/05/2006 11:03 AM, Stefan Manegold wrote with possible deletions: > Dear Peter, dear fellow PF & MXQ developers, > > I'm not sure, whether I do understand the new "ws-API" correctly, and would > be grateful if someone of you could enlighten me. > > First, here's how far I get so far (please correct me if/where I'm wrong!), > followed by concrete open questions. > > For transaction support, Peter introduced a new wsid that uniquely > identifies (a reference to?) a ws, which is required for conflict detection > (as far as I understand). > > A wsid is to be generated from/for a ws by calling wsid := ws_id(ws); > however, this generated a new unique id each time is called (two calls on > the same ws yield two different ids, right?), hence, wsid := ws_id(ws) > should immediately follow ws := ws_create() . > > For the inverse, there is ws := ws_bat(wsid) that return the ws identified > by wsid; obviously, this function yield the same result each time it's > called with the same wsid. > > Since some functionality required to to the wsid instead of "only" the ws, > Peter had changes the signatures/API of some runtime PROCs. As far as I see > right now, these are mainly (only?) > ws_doc(ws, ... -> ws_opendoc(wsid, ... > ws_addcoll(ws, ... -> ws_opencoll(wsid, ... > ws_destroy(int(ws)) -> ws_destroy(wsid) > > I finished these changes by simply running the folloing one-liners: > > find * -type f | xargs grep -l 'var ws := ws_create();' | xargs perl -i -p -e 's|(var ws := ws_create;)|$1 var wsid := ws_id(ws);|' > find * -type f | xargs grep -l 'ws_doc(ws,' | xargs perl -i -p -e 's|ws_doc$ws,|ws_opendoc(wsid,|' > find * -type f | xargs grep -l 'ws_destroy(int(ws))' | xargs perl -i -p -e 's|ws_destroy\(int\(ws$\)|ws_destroy(wsid)|' > find * -type f | xargs grep -l 'ws_addcoll(ws,' | xargs perl -i -p -e 's|ws_addcoll\(ws,|ws_opencoll(wsid,|' > find * -type f | xargs grep -l 'ws_addcoll' | xargs perl -i -p -e 's|ws_addcoll|ws_opencoll|' > > > While this seemed to have been enough to fix all .milS tests and (most of) > the milprint_summer functionallity, there are two open issues left: > the Algebra version, XRpc & PFtijah. > > To be honest, I haven't looked at PFtijah, yet; some help by the respective > experst would be appreciated. > > > In cases of XRpc & Algebra, there are still places were wsid := ws_id(ws) is > IMHO called in the wrong location: > > XRpc: > In PROC rpc_client() (runtime/xrpc.mx), I had changed > ws_addcoll(ws, ... > to > ws_opencoll(ws_id(ws), ... > While this seems to worg for now with read-only documents (i.e., in the > absense of updates and concurrency), it is IMHO incorrect in general. > Rather, rpc_client should receive wsid as argument instead of ws; > if necessary, ws can then be derived via ws := ws_bat(wsid). > > Algebra: > in PROC doc_tbl() (runtime/pf_support.mx), I had changed > var r := ws_doc(ws, item); > into > var wsid := ws_id(ws); > var r := ws_opendoc(wsid, item); > Same story as above. > However, doc_tbl return ws also in its result BAT-of-BATs, as far as I > understand mainly for "canonical API" reasons. > Obviously, it is not possible to replace BAT ws by lng wsid, here... > > > Leaving the latter problem aside for the moment, the following could (have) > work(ed) as a general rule, and we should check and change the codebase > accordingly: > > 1) wsid := ws_id(ws) must only be called immediately after ws := ws_create() > > 2) all functions/PROCs that recursively (i.e., including all transitively > called functions/PROCs) require wsid instead of or inaddition to ws should > be modified in that the receive wsid instead of ws as an argument; > ws can then locally be derived via ws := ws_bat(wsid) if/where necessary. > > 3) all functions/PROCs that recursively (i.e., including all transitively > called functions/PROCs) "never" (at least not yet) require wsid, but are > fine with ws can stay unchanged, i.e., getting (only) ws as argument. > Obviously, wherever these functions need to maintain a signature/API > aligned with those that fall under point 2) above, we should also change > these functions as described above with 2). > > > As indicated above, this does not work for doc_tbl and its companion PROCs > in runtime/pf_support.mx that make-up the interface between the Algebra > version of the compiler and the runtime. > I assume the respective canonical API could be changed in that the > respective PROC get wsid instead of (or in addition to?) ws as arguments. > However, we cannot easily change the API to return wsid (lng) instead of ws > (BAT) in the result BAT-of-BATs. Here, I see three solutions --- the last > one actually comes from JanR: > > a) Is ws/wsid indeed required in the result? > If not, we could simply discard it. > > b) Wrap wsid temprarly into a ("fake") [void,lng] BAT containing only one > (nil,wsid) BUN. > (Ugly!) > > c) (JanR) Put the wsid inside ws --- basically as a ("fake") BAT [void,lng]. > Kind of the "encapsulated" solution --- not only for the algebra-runtime > interface! > Consequently, we should have ws_create() call ws_id(ws) internally and > stored the result in ws (this could save the seemingly "redundant" > ws := ws_create(); wsid := ws_id(ws); sequence), and all (most, except at > least ws_destroy()) signatures/API could be kept / re-unified to pass > (only) ws (now including wsid) instead of wsid as argument. > > There seems to be one problem, so, if I understand Peter's comment > correctly: > " > - ws-IDs are now lng-s (combination of *unique* ID and bat-ID) such IDs > are meaningful also after the query is done (and ws-bat deallocated). > ^1^1^1^1^1^1^1^1^1^1^1^1^1^1 ^2^2^2^2^2^2^2^2^2^2^2 > this is needed for trans mgmt. > All meta-bats witha ws-id field now hold lng instead of int. > " > > In other words, (^2) wsid needs to be available also if ws is gone. > I'm not sure, though, whether is has to be available in the global wsid > variable (i.e., inside/during the query), or (only) in some meta-bats > outside (and hence after) the query (-context) (as suggested by ^1). > > > Peter, > > could you please enlighten me/us, here? > > > In case "only" (^1) is required, the "encapsulated" solution seems indeed > to be suitable for all code. > Otherwise, is would at least be a local solution for the algebra-runtime > interface (in case (a) & (b0 are no option), but bears the burden of > maintaining consistency between the global wsid variable and the wsid stored > inside the ws. > > > I hope, this email and any reactions to it help to clearify the current > situation, rather than to blur it even more ... > > Stefan > > -- Jan Rittinger Database Systems Technische Universit?t M?nchen (Germany) http://www-db.in.tum.de/~rittinge/ ------------------------------ Message: 4 Date: Fri, 6 Oct 2006 15:08:00 +0200 From: Stefan Manegold <Stefan.Manegold(a)cwi.nl> Subject: Re: [Monetdb-developers] The new "ws-API", Algebra & XRpc (& PFtijah) To: Jan Rittinger <rittinge(a)in.tum.de> Cc: Monetdb-developers(a)lists.sourceforge.net Message-ID: <20061006130800.GB5598(a)corona.ins.cwi.nl> Content-Type: text/plain; charset=us-ascii On Fri, Oct 06, 2006 at 02:39:19PM +0200, Jan Rittinger wrote: > If this wsid is a concept that is strictly necessary for the update > support (and nobody except me thinks that this bit shifting might be not > the nicest solution) we could also think about getting rid of the ws and > replacing it with the wsid. I just do not like the fact that there are > two variables underway that have some side effects on each other. To be honest, I'm still in the process of analyzing trying to understand, what all the flavors, pros, cons, and ideas behind having/using wsid are; hence, I cannot tell, whether it is "strictly necessary" or not. As far as I understand so far, the basic idea behind the "bit shifting" is merely to store both the id of (and hence reference to) BAT ws as well as a unique identifier --- that is still valid after a query/transaction (and hence cannot simply be the ws BAT id, which might be re-used for a subsequent query/transaction) --- in one single atomic value. Further, I agree, that having two related variables that need to be kept in sync and treated accordingly is not very handy for maintenance --- my guess is, that the current situation resulted from lack of time and fear to implement (something like) your (d) proposal; which I agree is "nicer" & "cleaner", but requires quite a lot of possibly error-prone code changes --- nevertheless, I might give it a try during the weekend ... Stefan > So a fourth proposal (d) would be to replace the occurrences of ws by > wsid. This would mean that *all* document accesses would have to include > another indirection (even all path steps). -- | Dr. Stefan Manegold | mailto:Stefan.Manegold@cwi.nl | | CWI, P.O.Box 94079 | http://www.cwi.nl/~manegold/ | | 1090 GB Amsterdam | Tel.: +31 (20) 592-4212 | | The Netherlands | Fax : +31 (20) 592-4312 | ------------------------------ Message: 5 Date: Fri, 6 Oct 2006 15:42:02 +0200 From: Stefan Manegold <Stefan.Manegold(a)cwi.nl> Subject: Re: [Monetdb-developers] The new "ws-API", Algebra & XRpc (& PFtijah) To: Monetdb-developers(a)lists.sourceforge.net Cc: Jan Rittinger <rittinge(a)in.tum.de> Message-ID: <20061006134202.GA6958(a)corona.ins.cwi.nl> Content-Type: text/plain; charset=us-ascii On Fri, Oct 06, 2006 at 03:08:00PM +0200, Stefan Manegold wrote: > On Fri, Oct 06, 2006 at 02:39:19PM +0200, Jan Rittinger wrote: > > If this wsid is a concept that is strictly necessary for the update > > support (and nobody except me thinks that this bit shifting might be not > > the nicest solution) we could also think about getting rid of the ws and > > replacing it with the wsid. I just do not like the fact that there are > > two variables underway that have some side effects on each other. > > To be honest, I'm still in the process of analyzing trying to understand, > what all the flavors, pros, cons, and ideas behind having/using wsid are; > hence, I cannot tell, whether it is "strictly necessary" or not. > > As far as I understand so far, the basic idea behind the "bit shifting" is > merely to store both the id of (and hence reference to) BAT ws as well as a > unique identifier --- that is still valid after a query/transaction (and hence > cannot simply be the ws BAT id, which might be re-used for a subsequent > query/transaction) --- in one single atomic value. > > Further, I agree, that having two related variables that need to be kept in > sync and treated accordingly is not very handy for maintenance --- my guess > is, that the current situation resulted from lack of time and fear to > implement (something like) your (d) proposal; which I agree is "nicer" & > "cleaner", but requires quite a lot of possibly error-prone code changes --- > nevertheless, I might give it a try during the weekend ... > > Stefan > > > So a fourth proposal (d) would be to replace the occurrences of ws by > > wsid. This would mean that *all* document accesses would have to include > > another indirection (even all path steps). Well, I just realize that this actually only helps to save us from having function interfaces that expect/use either ws or wsid or both --- it does not help to get rid of the global ws variable: to exist throught the whole query, the ws BAT must either be a global variable or be persistent --- the latter is not possible, since ws is a BAT-of-BATs, and those cannot be persistent... ... I just recall, that we have something like "session" BATs --- I'll check whether that's an option (any help/hint is of course welcome ...) Stefan -- | Dr. Stefan Manegold | mailto:Stefan.Manegold@cwi.nl | | CWI, P.O.Box 94079 | http://www.cwi.nl/~manegold/ | | 1090 GB Amsterdam | Tel.: +31 (20) 592-4212 | | The Netherlands | Fax : +31 (20) 592-4312 | ------------------------------ ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys -- and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV ------------------------------ _______________________________________________ Monetdb-developers mailing list Monetdb-developers(a)lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/monetdb-developers End of Monetdb-developers Digest, Vol 5, Issue 6 ************************************************

2 3

[Monetdb-developers] The new "ws-API", Algebra & XRpc (& PFtijah)
by Stefan Manegold 06 Oct '06

06 Oct '06

Dear Peter, dear fellow PF & MXQ developers, I'm not sure, whether I do understand the new "ws-API" correctly, and would be grateful if someone of you could enlighten me. First, here's how far I get so far (please correct me if/where I'm wrong!), followed by concrete open questions. For transaction support, Peter introduced a new wsid that uniquely identifies (a reference to?) a ws, which is required for conflict detection (as far as I understand). A wsid is to be generated from/for a ws by calling wsid := ws_id(ws); however, this generated a new unique id each time is called (two calls on the same ws yield two different ids, right?), hence, wsid := ws_id(ws) should immediately follow ws := ws_create() . For the inverse, there is ws := ws_bat(wsid) that return the ws identified by wsid; obviously, this function yield the same result each time it's called with the same wsid. Since some functionality required to to the wsid instead of "only" the ws, Peter had changes the signatures/API of some runtime PROCs. As far as I see right now, these are mainly (only?) ws_doc(ws, ... -> ws_opendoc(wsid, ... ws_addcoll(ws, ... -> ws_opencoll(wsid, ... ws_destroy(int(ws)) -> ws_destroy(wsid) I finished these changes by simply running the folloing one-liners: find * -type f | xargs grep -l 'var ws := ws_create();' | xargs perl -i -p -e 's|(var ws := ws_create;)|$1 var wsid := ws_id(ws);|' find * -type f | xargs grep -l 'ws_doc(ws,' | xargs perl -i -p -e 's|ws_doc$ws,|ws_opendoc(wsid,|' find * -type f | xargs grep -l 'ws_destroy(int(ws))' | xargs perl -i -p -e 's|ws_destroy\(int\(ws$\)|ws_destroy(wsid)|' find * -type f | xargs grep -l 'ws_addcoll(ws,' | xargs perl -i -p -e 's|ws_addcoll\(ws,|ws_opencoll(wsid,|' find * -type f | xargs grep -l 'ws_addcoll' | xargs perl -i -p -e 's|ws_addcoll|ws_opencoll|' While this seemed to have been enough to fix all .milS tests and (most of) the milprint_summer functionallity, there are two open issues left: the Algebra version, XRpc & PFtijah. To be honest, I haven't looked at PFtijah, yet; some help by the respective experst would be appreciated. In cases of XRpc & Algebra, there are still places were wsid := ws_id(ws) is IMHO called in the wrong location: XRpc: In PROC rpc_client() (runtime/xrpc.mx), I had changed ws_addcoll(ws, ... to ws_opencoll(ws_id(ws), ... While this seems to worg for now with read-only documents (i.e., in the absense of updates and concurrency), it is IMHO incorrect in general. Rather, rpc_client should receive wsid as argument instead of ws; if necessary, ws can then be derived via ws := ws_bat(wsid). Algebra: in PROC doc_tbl() (runtime/pf_support.mx), I had changed var r := ws_doc(ws, item); into var wsid := ws_id(ws); var r := ws_opendoc(wsid, item); Same story as above. However, doc_tbl return ws also in its result BAT-of-BATs, as far as I understand mainly for "canonical API" reasons. Obviously, it is not possible to replace BAT ws by lng wsid, here... Leaving the latter problem aside for the moment, the following could (have) work(ed) as a general rule, and we should check and change the codebase accordingly: 1) wsid := ws_id(ws) must only be called immediately after ws := ws_create() 2) all functions/PROCs that recursively (i.e., including all transitively called functions/PROCs) require wsid instead of or inaddition to ws should be modified in that the receive wsid instead of ws as an argument; ws can then locally be derived via ws := ws_bat(wsid) if/where necessary. 3) all functions/PROCs that recursively (i.e., including all transitively called functions/PROCs) "never" (at least not yet) require wsid, but are fine with ws can stay unchanged, i.e., getting (only) ws as argument. Obviously, wherever these functions need to maintain a signature/API aligned with those that fall under point 2) above, we should also change these functions as described above with 2). As indicated above, this does not work for doc_tbl and its companion PROCs in runtime/pf_support.mx that make-up the interface between the Algebra version of the compiler and the runtime. I assume the respective canonical API could be changed in that the respective PROC get wsid instead of (or in addition to?) ws as arguments. However, we cannot easily change the API to return wsid (lng) instead of ws (BAT) in the result BAT-of-BATs. Here, I see three solutions --- the last one actually comes from JanR: a) Is ws/wsid indeed required in the result? If not, we could simply discard it. b) Wrap wsid temprarly into a ("fake") [void,lng] BAT containing only one (nil,wsid) BUN. (Ugly!) c) (JanR) Put the wsid inside ws --- basically as a ("fake") BAT [void,lng]. Kind of the "encapsulated" solution --- not only for the algebra-runtime interface! Consequently, we should have ws_create() call ws_id(ws) internally and stored the result in ws (this could save the seemingly "redundant" ws := ws_create(); wsid := ws_id(ws); sequence), and all (most, except at least ws_destroy()) signatures/API could be kept / re-unified to pass (only) ws (now including wsid) instead of wsid as argument. There seems to be one problem, so, if I understand Peter's comment correctly: " - ws-IDs are now lng-s (combination of *unique* ID and bat-ID) such IDs are meaningful also after the query is done (and ws-bat deallocated). ^1^1^1^1^1^1^1^1^1^1^1^1^1^1 ^2^2^2^2^2^2^2^2^2^2^2 this is needed for trans mgmt. All meta-bats witha ws-id field now hold lng instead of int. " In other words, (^2) wsid needs to be available also if ws is gone. I'm not sure, though, whether is has to be available in the global wsid variable (i.e., inside/during the query), or (only) in some meta-bats outside (and hence after) the query (-context) (as suggested by ^1). Peter, could you please enlighten me/us, here? In case "only" (^1) is required, the "encapsulated" solution seems indeed to be suitable for all code. Otherwise, is would at least be a local solution for the algebra-runtime interface (in case (a) & (b0 are no option), but bears the burden of maintaining consistency between the global wsid variable and the wsid stored inside the ws. I hope, this email and any reactions to it help to clearify the current situation, rather than to blur it even more ... Stefan -- | Dr. Stefan Manegold | mailto:Stefan.Manegold@cwi.nl | | CWI, P.O.Box 94079 | http://www.cwi.nl/~manegold/ | | 1090 GB Amsterdam | Tel.: +31 (20) 592-4212 | | The Netherlands | Fax : +31 (20) 592-4312 |

2 4

[Monetdb-developers] bolts damaged
by server RTz 04 Oct '06

04 Oct '06

attack applies contact area ordinary wrenches Kroc. HSW Fact Day: paintball markers Foresters Boards Blogs Segments SNAP High Candy Spiked same Guarana fuels amazing BAWLS drink Brown customers saying... Copyright Rights Reserved Statement Company Music.com Reviews Videos support. Resources Getting v. Edition. In earlier releases used similar OS v. Edition. In earlier releases used similar date: Copyright Rights Reserved Statement Company Music.com Reviews Videos InNew User Connect S.com Market Forums Boards Blogs Segments SNAP Next Gaming PRO SAP Oracle Music Palm BREW Cross iMode Windows UIQ Knowledge Known Issues FAQs Tips eUpdate archive Platforms Military Physical drives. gigs Ionut Ambrosie winners meet Monitor home Travel Consumer Products Bawls Mints High Candy

1 0

[Monetdb-developers] Error compiling MEL w/ flex 2.5.33 & bison 2.3
by Stefan Manegold 03 Oct '06

03 Oct '06

On Mon, Oct 02, 2006 at 12:47:39PM +0200, Niels Nes wrote: > On Mon, Oct 02, 2006 at 12:32:22PM +0200, Riham wrote: > > Hi, > > > > I am a new PhD student at the University of Twente. I am working with > > Dr. Maurice van Keulen on Optimizing Pathfinder queries. I was > > installing Monet on my new linux system and I got the following error: > > > > =========================================================================== > > kader@netloosewi840:~/home/MonetDB_XQuery_CVS/build/buildtools> make > > cd > > ""/local/kader/home/MonetDB_XQuery_CVS/build/buildtools/../../src/buildtools"/autogen" > > && python setup.py build > > running build > > running build_py > > running build_scripts > > cd _MX && make MAKE=make all > > make[1]: Entering directory > > `/local/kader/home/MonetDB_XQuery_CVS/build/buildtools/_MX' > > make all-am > > make[2]: Entering directory > > `/local/kader/home/MonetDB_XQuery_CVS/build/buildtools/_MX' > > make[2]: Nothing to be done for `all-am'. > > make[2]: Leaving directory > > `/local/kader/home/MonetDB_XQuery_CVS/build/buildtools/_MX' > > make[1]: Leaving directory > > `/local/kader/home/MonetDB_XQuery_CVS/build/buildtools/_MX' > > cd _MEL && make MAKE=make all > > make[1]: Entering directory > > `/local/kader/home/MonetDB_XQuery_CVS/build/buildtools/_MEL' > > make all-am > > make[2]: Entering directory > > `/local/kader/home/MonetDB_XQuery_CVS/build/buildtools/_MEL' > > g++ -g -o mel any_arg.o atom_arg.o atom.o atomops.o bat_arg.o > > builtin.o command.o debug.o depend.o dependency.o epilogue.o FileInfo.o > > glue.o html.o iterator.o link.o ListIterator.o mel.o mel.tab.o mel.yy.o > > mil.o module.o object.o operator.o ops.o prelude.o proto.o symbol.o > > symtable.o type_arg.o use.o var_arg.o > > mel.tab.o: In function `semerror()': > > /local/kader/home/MonetDB_XQuery_CVS/build/buildtools/_MEL/mel.yy:471: > > undefined reference to `yylineno' > > /local/kader/home/MonetDB_XQuery_CVS/build/buildtools/_MEL/mel.yy:472: > > undefined reference to `yylineno' > > mel.tab.o: In function `yyparse()': > > /local/kader/home/MonetDB_XQuery_CVS/build/buildtools/_MEL/mel.yy:59: > > undefined reference to `yylineno' > > /local/kader/home/MonetDB_XQuery_CVS/build/buildtools/_MEL/mel.yy:67: > > undefined reference to `yylineno' > > /local/kader/home/MonetDB_XQuery_CVS/build/buildtools/_MEL/mel.yy:131: > > undefined reference to `yylineno' > > mel.yy.o:/local/kader/home/MonetDB_XQuery_CVS/build/buildtools/_MEL/mel.ll:253: > > more undefined references to `yylineno' follow > > collect2: ld returned 1 exit status > > make[2]: *** [mel] Error 1 > > make[2]: Leaving directory > > `/local/kader/home/MonetDB_XQuery_CVS/build/buildtools/_MEL' > > make[1]: *** [all] Error 2 > > make[1]: Leaving directory > > `/local/kader/home/MonetDB_XQuery_CVS/build/buildtools/_MEL' > > make: *** [all] Error 2 > > =========================================================================== > > > > The flex and bison versions I have are 2.5.33 and 2.3 respectively. > > I was advised by Jan Flokstra to report the error to you. I hope you > > have sufficient information to know the origin of the problem, if not > > please let me know. > > > > Thank you for your help, > > Riham Abdel Kader. > > As of flex 2.5.31 the yylineno was defined by flex. Normaly we define > it our selfs. > > You could try to patch the following line from mel.mx > > #if !defined(YY_FLEX_SUBMINOR_VERSION) || YY_FLEX_SUBMINOR_VERSION < 5 > /* flex 2.5.31 declares yylineno. I don't know which other versions > do as well. */ > int yylineno; > #endif > > Simply removed the #if/#endif should work. A better solution is to know > which range between 2.5.6 and 2.5.31/32 have a yylineno defined. And > fix the second part of the #ifdef. > > Niels Niels, Riham, I haven't analysed the problem any further, yet, but I think it might also be related to bison 2.3 or the combination of bison 2.3 and flex 2.5.33. In out nightly testing pool (cf., http://monetdb.cwi.nl/Development/TestWeb/) we use flex 2.5.4 & flex 2.5.33 and bison 1.875, 2.0, 2.1, 2.2, and everything works fine. We don't use bison 2.3, yet ... Riham, which linux distribution (name & version) are you using? Does Niels' patch work for you? Stefan -- | Dr. Stefan Manegold | mailto:Stefan.Manegold@cwi.nl | | CWI, P.O.Box 94079 | http://www.cwi.nl/~manegold/ | | 1090 GB Amsterdam | Tel.: +31 (20) 592-4212 | | The Netherlands | Fax : +31 (20) 592-4312 |

4 11