Menu
▾
▴
firebird-devel
|
From: Leyne, S. <sl...@at...> - 2001-05-07 13:33:55
|
Mike,
How about just calling the directory "generated"?
Sean
-----Original Message-----
From: Mike Nordell [mailto:ta...@al...]
Sent: Monday, May 07, 2001 9:17 AM
To: fir...@li...
Subject: [Firebird-devel] Macros, autoconf and Firebird 2 [was
Bugs-420205 ]
To give a taste of the hierarchy that is waiting:
<fb_root>/
src/
builds/
builds_win32/
<generated_src>/ (.e -> .c)
examples/
<generated_src> is currently called "generated_src", if anyone has a
better
idea...
The reason for this mirror hierarchy is to have a "clean" src/ tree - a
plain CVS mirror with no extra files.
|
|
From: Mike N. <ta...@al...> - 2001-05-09 15:08:28
|
Sorry for the late response. It got lost in the flood of posts. Leyne, Sean wrote: > How about just calling the directory "generated"? Sure. /Mike |
|
From: Ann W. H. <aha...@ib...> - 2001-05-09 15:58:08
|
I've been looking at a problem 'til I'm cross-eyed and it's time for some help. My copy of Firebird and earlier versions of both Firebird and InterBase get a "wrong length record (183)" error on the second or third record in the RDB$FUNCTIONS table of one particular V5.6 database. The certified InterBase does not, though I can't find any changes on sourceforge that might account for the difference. Nor does inspecting the record suggest what might be wrong with it (other than typical harrison muddleheadedness). There is a database available as a zip file from ftp at firebird.netfrastructure.com - anonymous login, get the filed called e1gdb.zip. It should work on either Windows or Linux. It's an ODS 9 database, of course. Here are things I would like: Could someone who has a built copy of InterBase Open Edition try to backup this database using that engine? Could someone who has a recent Firebird build (blush) try the same? Could someone figure out what's going on? Regards, Ann The rest of this might be of interest to people thinking about my third request. Here's some more detailed information: The message comes from VIO.c, but is likely to have been caused by an error in SQZ.c. This is the routine in VIO. For those who don't know, a tdbb is a thread specific database block, a pool is an area where space can be allocated for a specific function and an rpb is a record parameter block. Once a record has been lifted off page, it travels around in an rpb. Active and inactive refer to whether the page with the record is locked. The record in question is not a back version, or a fragmented record, so this code should execute it's simplest path - essentially just getting the bits and decompressing them. void VIO_data ( TDBB tdbb, register RPB *rpb, BLK pool) { /************************************** * * V I O _ d a t a * ************************************** * * Functional description * Given an active record parameter block, fetch the full record. * * This routine is called with an active RPB and exits with * an INactive RPB. Yes, Virginia, getting the data for a * record means losing control of the record. This turns out * to matter a lot. **************************************/ FMT format; REC record, prior; ULONG back_page; USHORT length, back_line; UCHAR *tail, *tail_end, differences [MAX_DIFFERENCES]; SET_TDBB (tdbb); /* If we're not already set up for this format version number, find the format block and set up the record block. This is a performance optimization. */ record = VIO_record (tdbb, rpb, NULL_PTR, pool); format = record->rec_format; /* If the record is a delta version, start with data from prior record. */ prior = rpb->rpb_prior; if (prior) { tail = differences; tail_end = differences + sizeof (differences); if (prior != record) MOVE_FASTER (prior->rec_data, record->rec_data, format->fmt_length); } else { tail = record->rec_data; tail_end = tail + record->rec_length; } /* Set up prior record point for next version */ rpb->rpb_prior = (rpb->rpb_b_page && (rpb->rpb_flags & rpb_delta)) ? record : 0; /* Snarf data from record */ tail = SQZ_decompress (rpb->rpb_address, rpb->rpb_length, tail, tail_end); if (rpb->rpb_flags & rpb_incomplete) { back_page = rpb->rpb_b_page; back_line = rpb->rpb_b_line; while (rpb->rpb_flags & rpb_incomplete) { DPM_fetch_fragment (tdbb, rpb, LCK_read); tail = SQZ_decompress (rpb->rpb_address, rpb->rpb_length, tail, tail_end); } rpb->rpb_b_page = back_page; rpb->rpb_b_line = back_line; } /* unlock the page */ CCH_RELEASE (tdbb, &rpb->rpb_window); /* If this is a delta version, apply changes */ if (prior) length = SQZ_apply_differences (record, differences, tail); else length = tail - record->rec_data; if (format->fmt_length != length) BUGCHECK (183); /* msg 183 wrong record length */ rpb->rpb_address = record->rec_data; rpb->rpb_length = format->fmt_length; } The problem, more than likely, is in SQZ - squeeze. Here's SQZ decompress. As you know, compression is a run length encoding using a byte length. The values from 1 - 127 indicate that the bytes that follow are to be copied directly to the output. The values from -3 to -128 indicate that the byte that follows is to be replicated (-1 times )that many times. SCHAR *SQZ_decompress ( SCHAR *input, USHORT length, SCHAR *output, SCHAR *output_end) { /************************************** * * S Q Z _ d e c o m p r e s s * ************************************** * * Functional description * Decompress a compressed string into a buffer. Return the address * where the output stopped. * **************************************/ SCHAR *last, c; SSHORT l; last = input + length; while (input < last) { l = *input++; if (l < 0) { c = *input++; if ((output -l) > output_end) BUGCHECK (179); /* msg 179 decompression overran buffer */ memset (output, (UCHAR) c, (-1*l) ); output -= l; } else { if ((output + l) > output_end) BUGCHECK (179); /* msg 179 decompression overran buffer */ MOVE_FAST(input,output,l); output +=l; input +=l; } } if (output > output_end) BUGCHECK (179); /* msg 179 decompression overran buffer */ return output; } And, as if this weren't enough, here is a dump of the page with the error. I believe that the problem is with the second record on page (1) but it may be with the third (2). My dump program may have problems... who doesn't? 214 DATA PAGE checksum 12345 generation 1 Relation: 14, Sequence: 0, Count: 26, Flags: 0 - offset: 4016, length: 79, expanded data length: 370 trans: 1, format: 0 Compressed format: (length 66) 01 80 fd 00 0e F_DEFAULTFLOATef fd 00 e1 f5 00 10 0e 00 Fre eUDFLib.dll80 00 91 00 0d Default_Floatee fc 00 Decompressed format: (length 370) 80 00 00 00 F_DEFAULTFLOAT 00 00 00 00 00 00 00 00 00 00 00 00 00 00 0e 00 FreeUDFLib.dll00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 Default_Float 00 00 00 00 1 - offset: 3924, length: 90, expanded data length: 370 trans: 1, format: 0 Compressed format: (length 77) 01 80 fd 00 14 F_AGEINDAYSTHRESHOLDf5 fd 00 e1 f5 00 10 0e 00 FreeUDFLib.dll80 00 91 00 12 AgeInDaysThresholdf3 fc 00 Decompressed format: (length 370) 80 00 00 00 F_AGEINDAYSTHRESHOLD 00 00 00 00 00 00 00 00 00 00 00 00 00 00 0e 00 FreeUDFLib.dll00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 AgeInDaysThreshold 00 00 00 00 2 - offset: 3860, length: 61, expanded data length: 370 trans: 1, format: 0 Compressed format: (length 48) 01 80 fd 00 05 FLOORe6 fd 00 e1 f5 00 08 06 00 ib_udf80 00 89 00 0c IB_UDF_floored fc 00 Decompressed format: (length 370) 80 00 00 00 FLOOR 00 00 00 00 00 00 00 00 00 00 00 00 00 00 06 00 ib_udf00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 IB_UDF_ floor 00 00 00 00 And finally, here's a dump of the same page, created by the V6 engine, which works perfectly. 214 DATA PAGE checksum 12345 generation 1 Relation: 14, Sequence: 0, Count: 26, Flags: 0 - offset: 4016, length: 79, expanded data length: 370 trans: 1, format: 0 Compressed format: (length 66) 01 80 fd 00 0e F_DEFAULTFLOATef fd 00 e1 f5 00 10 0e 00 Fre eUDFLib.dll80 00 91 00 0d Default_Floatee fc 00 Decompressed format: (length 370) 80 00 00 00 F_DEFAULTFLOAT 00 00 00 00 00 00 00 00 00 00 00 00 00 00 0e 00 FreeUDFLib.dll00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 Default_Float 00 00 00 00 1 - offset: 3924, length: 90, expanded data length: 370 trans: 1, format: 0 Compressed format: (length 77) 01 80 fd 00 14 F_AGEINDAYSTHRESHOLDf5 fd 00 e1 f5 00 10 0e 00 FreeUDFLib.dll80 00 91 00 12 AgeInDaysThresholdf3 fc 00 Decompressed format: (length 370) 80 00 00 00 F_AGEINDAYSTHRESHOLD 00 00 00 00 00 00 00 00 00 00 00 00 00 00 0e 00 FreeUDFLib.dll00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 AgeInDaysThreshold 00 00 00 00 2 - offset: 3860, length: 61, expanded data length: 370 trans: 1, format: 0 Compressed format: (length 48) 01 80 fd 00 05 FLOORe6 fd 00 e1 f5 00 08 06 00 ib_udf80 00 89 00 0c IB_UDF_floored fc 00 Decompressed format: (length 370) 80 00 00 00 FLOOR 00 00 00 00 00 00 00 00 00 00 00 00 00 00 06 00 ib_udf00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 IB_UDF_ floor 00 00 00 00 3 - offset: 3788, length: 72, expanded data length: 370 trans: 1, format: 0 Compressed format: (length 59) 01 80 fd 00 0a F_STRTOINTeb fd 00 e1 f5 00 10 0e 00 FreeUDF Lib.dll80 00 91 00 0a Str_To_Inteb fc 00 Decompressed format: (length 370) 80 00 00 00 F_STRTOINT 00 00 00 00 00 00 00 00 00 00 00 00 00 00 0e 00 FreeUDFLib.dll00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 Str_To_Int 00 00 00 00 Regards, Ann www.ibphoenix.com We have answers. |
|
From: Mike N. <ta...@al...> - 2001-05-10 04:53:43
|
Ann W. Harrison wrote:
> My copy of Firebird and earlier versions
> of both Firebird and InterBase get a "wrong length record (183)"
> error on the second or third record in the RDB$FUNCTIONS table
> of one particular V5.6 database.
First, do you know if this works using the original tarball, or does that
one also fail?
Try adding a static size_t and increment it everytime you enter VIO_data. It
will probably break when the counter reaches 7352. It gives you a nice
counter to break on in the debugger. Using Win32, if you're lazy, you can
even add
static size_t nEntries = 0;
++nEntries;
if (nEntries == 0x1cb8) {
__asm int 3;
}
What's wrong here is that he record format doesn't match the length of the
decompressed block. SQZ is fine, it's bad earlier.
The function it tries to write is probably "F_DEFAULTFLOAT" (last one
printed in the gbak console), while the particular relation is
"RDB$FUNCTION_ARGUMENTS".
From tracing into MET_format() (via VIO_record), it already here display
that this relation.rel_cur_format.fmt_length is 0x34. What we get from
SQZ_decompress later on is 0x32. It's bad input even earlier.
The compressed buffer that breaks, len 0x1f, looks like
01 80 FD 00 0E F_DEFAULTFLOAT EF 20 FB 00 01 1B FB 00 01 08 FB 00
The closest I get back in the call stack before becoming completely lost is
rse.c get_record() on the rsb_indexed case. Well, if you don't count how it
gets the RPB* in get_record():
request->req_rpb + rsb->rsb_stream
rsb->rsb_stream is zero, why it's the plain request->req_rpb that it gets.
> The certified InterBase does
> not, though I can't find any changes on sourceforge that might
> account for the difference.
Then we definitely have to dig deeper. Have you tried the *hrm* mers build
*hrm*?
> Could someone who has a recent Firebird build (blush) try
> the same?
It breaks.
> The values from -3 to -128 indicate that the byte
> that follows is to be replicated (-1 times )that many times.
Are -1 and -2 illegal values? If so, I think it should be asserted.
I even ran the server under Purify to rule out buffer overflows. No detected
problems.
Not much of a help, but I hope it can speed you on your way.
/Mike
|
|
From: <Sch...@t-...> - 2001-05-10 07:01:48
|
Mike Nordell wrote:
> > Could someone who has a recent Firebird build (blush) try
> > the same?
>
> It breaks.
I've tried it with my (classic server)
ISQL Version: LI-T0.9.5.145 Firebird Beta1
and
gbak: gbak version Firebird-W32-V1.0 (huh? what's this)
and it works, (restore is complaining about missing udf's; ok, I'm on
linux)
earlier versions choke with
internal gds software consistency check
(wrong record length (183))
when trying to acces eg. ARCBRANCH
hth
Frank
--
"Fascinating creatures, phoenixes. They can carry immensely heavy loads,
their tears have healing powers and they make highly faithful pets."
- J.K. Rowling
http://firebird.sourceforge.net
|
|
From: Ann W. H. <aha...@ib...> - 2001-05-10 13:48:58
|
At 06:53 AM 5/10/2001 +0200, Mike Nordell wrote: > > both Firebird and InterBase get a "wrong length record (183)" > >First, do you know if this works using the original tarball Fails >The function it tries to write is probably "F_DEFAULTFLOAT" (last one >printed in the gbak console), while the particular relation is >"RDB$FUNCTION_ARGUMENTS". Unh, duh! right. Looking in the wrong place. Silly me. > > The values from -3 to -128 indicate that the byte > > that follows is to be replicated (-1 times )that many times. > >Are -1 and -2 illegal values? If so, I think it should be asserted. O, -1, & -2 are not used. The thought has been bruited around to use one of those to indicate that the next two bytes are length. Something to do for V2. The problem it solves is that an empty 32K varchar currently uses about 300 bytes of storage. Regards, Ann who is feeling really dim. |
×
Want the latest updates on software, tech news, and AI?
Get latest updates about software, tech news, and AI from SourceForge directly in your inbox once a month.
Thanks for helping keep SourceForge clean.
X