Re: [HACKERS] compile bug in HEAD?
Neil Conway wrote: On Wed, Mar 27, 2002 at 07:56:15PM -0500, Peter Eisentraut wrote: Neil Conway writes: I'm curious; why is this not the right fix? According to the manpage: -lturns on maximum compatibility with the original ATT lex implementation. Note that this does not mean full compatibility. Use of this option costs a considerable amount of performance... The manpage also lists the specific incompatibilities. I think we should not be affected by them, but someone better check before removing the -l. AFAICT current sources don't actually use -l anywhere. However, it does appear that we can tweak flex for more performance (usually at the expense of a larger generated parser). In particular, it looks like we could use -Cf or -CF. Is this a good idea? While we're on the subject of minor optimizations, is there a reason why we execute gcc with -O2 rather than -O3 during compilation? Added to TODO: * Try flex flags -Cf and -CF to see if performance improves -- Bruce Momjian| http://candle.pha.pa.us [EMAIL PROTECTED] | (610) 853-3000 + If your life is a hard drive, | 830 Blythe Avenue + Christ can be your backup.| Drexel Hill, Pennsylvania 19026 ---(end of broadcast)--- TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]
Re: [HACKERS] compile bug in HEAD?
On Wed, Mar 27, 2002 at 07:56:15PM -0500, Peter Eisentraut wrote: Neil Conway writes: I'm curious; why is this not the right fix? According to the manpage: -l turns on maximum compatibility with the original ATT lex implementation. Note that this does not mean full compatibility. Use of this option costs a considerable amount of performance... The manpage also lists the specific incompatibilities. I think we should not be affected by them, but someone better check before removing the -l. AFAICT current sources don't actually use -l anywhere. However, it does appear that we can tweak flex for more performance (usually at the expense of a larger generated parser). In particular, it looks like we could use -Cf or -CF. Is this a good idea? While we're on the subject of minor optimizations, is there a reason why we execute gcc with -O2 rather than -O3 during compilation? Cheers, Neil -- Neil Conway [EMAIL PROTECTED] PGP Key ID: DB3C29FC ---(end of broadcast)--- TIP 6: Have you searched our list archives? http://archives.postgresql.org
Re: [HACKERS] compile bug in HEAD?
Neil Conway writes: However, it does appear that we can tweak flex for more performance (usually at the expense of a larger generated parser). In particular, it looks like we could use -Cf or -CF. Is this a good idea? Probably. Run some performance tests if you like. It looks like -CFea might be a reasonable candidate. While we're on the subject of minor optimizations, is there a reason why we execute gcc with -O2 rather than -O3 during compilation? Mainly because everyone does it this way. Probably because it's a reasonable compromise between execution speed, compilation speed, debuggability, and compiler bugs. -- Peter Eisentraut [EMAIL PROTECTED] ---(end of broadcast)--- TIP 3: if posting/reading through Usenet, please send an appropriate subscribe-nomail command to [EMAIL PROTECTED] so that your message can get through to the mailing list cleanly
Re: [HACKERS] compile bug in HEAD?
On Wed, 2002-03-27 at 11:06, Peter Eisentraut wrote: If someone is really bored out of their mind, at least one of these warnings can be gotten rid of by not using the -l option to flex. That might be desirable for other reasons, too, one of which is improved speed. No, just removing -l from the makefile is not the right fix. I'm curious; why is this not the right fix? According to the manpage: -l turns on maximum compatibility with the original ATT lex implementation. Note that this does not mean full compatibility. Use of this option costs a considerable amount of performance... Cheers, Neil -- Neil Conway [EMAIL PROTECTED] PGP Key ID: DB3C29FC ---(end of broadcast)--- TIP 2: you can get off all lists at once with the unregister command (send unregister YourEmailAddressHere to [EMAIL PROTECTED])
Re: [HACKERS] compile bug in HEAD?
Neil Conway writes: I'm curious; why is this not the right fix? According to the manpage: -lturns on maximum compatibility with the original ATT lex implementation. Note that this does not mean full compatibility. Use of this option costs a considerable amount of performance... The manpage also lists the specific incompatibilities. I think we should not be affected by them, but someone better check before removing the -l. -- Peter Eisentraut [EMAIL PROTECTED] ---(end of broadcast)--- TIP 2: you can get off all lists at once with the unregister command (send unregister YourEmailAddressHere to [EMAIL PROTECTED])
Re: [HACKERS] compile bug in HEAD?
It's more warnings than bugs. I also have seen that but not familiar enough with bison or yacc to think more of it. Have you got an idea on how to fix these warnings? - Original Message - From: Christopher Kings-Lynne [EMAIL PROTECTED] To: Hackers [EMAIL PROTECTED] Sent: Wednesday, March 27, 2002 4:24 PM Subject: [HACKERS] compile bug in HEAD? I don't think this is me... gcc -pipe -O -Wall -Wmissing-prototypes -Wmissing-declarations -Wno-error -I ./../include -I. -I../../../../src/include -DMAJOR_VERSION=2 -DMINOR_VERSIO N=10 -DPATCHLEVEL=0 -DINCLUDE_PATH=\/home/chriskl/local/include\ -c -o pgc.o pgc.c pgc.c: In function `yylex': pgc.c:1250: warning: label `find_rule' defined but not used pgc.l: At top level: pgc.c:3079: warning: `yy_flex_realloc' defined but not used Chris ---(end of broadcast)--- TIP 5: Have you checked our extensive FAQ? http://www.postgresql.org/users-lounge/docs/faq.html ---(end of broadcast)--- TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]
Re: [HACKERS] compile bug in HEAD?
Christopher Kings-Lynne wrote: I don't think this is me... gcc -pipe -O -Wall -Wmissing-prototypes -Wmissing-declarations -Wno-error -I ./../include -I. -I../../../../src/include -DMAJOR_VERSION=2 -DMINOR_VERSIO N=10 -DPATCHLEVEL=0 -DINCLUDE_PATH=\/home/chriskl/local/include\ -c -o pgc.o pgc.c pgc.c: In function `yylex': pgc.c:1250: warning: label `find_rule' defined but not used pgc.l: At top level: pgc.c:3079: warning: `yy_flex_realloc' defined but not used Yes, I have gotten the same warning for several releases but haven't researched the cause. Patch? -- Bruce Momjian| http://candle.pha.pa.us [EMAIL PROTECTED] | (610) 853-3000 + If your life is a hard drive, | 830 Blythe Avenue + Christ can be your backup.| Drexel Hill, Pennsylvania 19026 ---(end of broadcast)--- TIP 2: you can get off all lists at once with the unregister command (send unregister YourEmailAddressHere to [EMAIL PROTECTED])
Re: [HACKERS] compile bug in HEAD?
Nicolas Bazin [EMAIL PROTECTED] writes: It's more warnings than bugs. I also have seen that but not familiar enough with bison or yacc to think more of it. Have you got an idea on how to fix these warnings? ecpg's lexer has always generated those warnings, and so has plpgsql's lexer. AFAICT the sloppy C code is triggered by use of yylineno. Suggest griping to the flex authors. regards, tom lane ---(end of broadcast)--- TIP 3: if posting/reading through Usenet, please send an appropriate subscribe-nomail command to [EMAIL PROTECTED] so that your message can get through to the mailing list cleanly