[PHP-DEV] Re: [PHP-CVS] cvs: php4 /ext/standard/tests/file userstreams.phpt
On 10/29/02, Marcus B?rger [EMAIL PROTECTED] wrote: -Use // instead of /* ... */ comments Why? #this test fails for me. Very weired... You changed the test. Why not ask me first about it failing? --Wez. -- PHP Development Mailing List http://www.php.net/ To unsubscribe, visit: http://www.php.net/unsub.php
[PHP-DEV] Re: [PHP-CVS] cvs: php4 /ext/standard/tests/file userstreams.phpt
I started to work on that issue. My first attempt was to check where and why the test failed. I used the comment modification because it seems the test fails due to some internal failure. The reason i did not ask you before was that i considered my changes a kind of whitespace change so that any further change can easily be seen. The question i would like to ask now we already talk about this is: When i comment out everythink starting at ob_start() i get some output. If not (original) then i don't get any output. What is weired is that when i use any print/echo statement after ob_end_xxx() nothing is printed... The length of the lyrics added by me was only to test whether or not that part of the script works. marcus At 15:25 29.10.2002, Wez Furlong wrote: On 10/29/02, Marcus B?rger [EMAIL PROTECTED] wrote: -Use // instead of /* ... */ comments Why? #this test fails for me. Very weired... You changed the test. Why not ask me first about it failing? --Wez. -- PHP Development Mailing List http://www.php.net/ To unsubscribe, visit: http://www.php.net/unsub.php
[PHP-DEV] Re: [PHP-CVS] cvs: php4 /ext/standard/tests/file userstreams.phpt
Lets stick to C comments, as those are the preference listed in coding standards (and C++ style comments are really only intended for single line out-commenting anyhow). The test works fine here; it relies on output buffering working correctly; when the test passes you see only the expected output. If it fails, you see a trace of the seek offsets and expected positioning etc. etc. The failure of the test for you, and also the strange behaviour when changing the ob_xxx calls points a finger at the output buffer layer. Lets just make it a point to discuss something before changing someone elses code: comments and coding styles are often a touchy subject with a lot of developers (I'm normally fine with it, but it seemed pointless in this case, and against our general coding standards.) --Wez. On 10/29/02, Marcus Börger [EMAIL PROTECTED] wrote: I started to work on that issue. My first attempt was to check where and why the test failed. I used the comment modification because it seems the test fails due to some internal failure. The reason i did not ask you before was that i considered my changes a kind of whitespace change so that any further change can easily be seen. The question i would like to ask now we already talk about this is: When i comment out everythink starting at ob_start() i get some output. If not (original) then i don't get any output. What is weired is that when i use any print/echo statement after ob_end_xxx() nothing is printed... The length of the lyrics added by me was only to test whether or not that part of the script works. marcus At 15:25 29.10.2002, Wez Furlong wrote: On 10/29/02, Marcus B?rger [EMAIL PROTECTED] wrote: -Use // instead of /* ... */ comments Why? #this test fails for me. Very weired... You changed the test. Why not ask me first about it failing? --Wez. -- PHP Development Mailing List http://www.php.net/ To unsubscribe, visit: http://www.php.net/unsub.php
Re: [PHP-DEV] Re: [PHP-CVS] cvs: php4 /ext/standard/tests/file userstreams.phpt
This test fails with adding --INI-- output_buffering=0 output_handler= zlib.output_compression= implicit_flush=0 and even when removing the three calls to ob_xxx89 At 15:59 29.10.2002, Wez Furlong wrote: Lets stick to C comments, as those are the preference listed in coding standards (and C++ style comments are really only intended for single line out-commenting anyhow). The test works fine here; it relies on output buffering working correctly; when the test passes you see only the expected output. If it fails, you see a trace of the seek offsets and expected positioning etc. etc. The failure of the test for you, and also the strange behaviour when changing the ob_xxx calls points a finger at the output buffer layer. Lets just make it a point to discuss something before changing someone elses code: comments and coding styles are often a touchy subject with a lot of developers (I'm normally fine with it, but it seemed pointless in this case, and against our general coding standards.) --Wez. On 10/29/02, Marcus Börger [EMAIL PROTECTED] wrote: I started to work on that issue. My first attempt was to check where and why the test failed. I used the comment modification because it seems the test fails due to some internal failure. The reason i did not ask you before was that i considered my changes a kind of whitespace change so that any further change can easily be seen. The question i would like to ask now we already talk about this is: When i comment out everythink starting at ob_start() i get some output. If not (original) then i don't get any output. What is weired is that when i use any print/echo statement after ob_end_xxx() nothing is printed... The length of the lyrics added by me was only to test whether or not that part of the script works. marcus At 15:25 29.10.2002, Wez Furlong wrote: On 10/29/02, Marcus B?rger [EMAIL PROTECTED] wrote: -Use // instead of /* ... */ comments Why? #this test fails for me. Very weired... You changed the test. Why not ask me first about it failing? --Wez. -- PHP Development Mailing List http://www.php.net/ To unsubscribe, visit: http://www.php.net/unsub.php -- PHP Development Mailing List http://www.php.net/ To unsubscribe, visit: http://www.php.net/unsub.php
Re: [PHP-DEV] Re: [PHP-CVS] cvs: php4 /ext/standard/tests/file userstreams.phpt
On 10/29/02, Marcus Börger [EMAIL PROTECTED] wrote: This test fails with adding --INI-- output_buffering=0 output_handler= zlib.output_compression= implicit_flush=0 Works fine here. and even when removing the three calls to ob_xxx89 Can you email me the output? Also, could you strace it too? (just to be sure) --Wez. -- PHP Development Mailing List http://www.php.net/ To unsubscribe, visit: http://www.php.net/unsub.php
Re: [PHP-DEV] Re: [PHP-CVS] cvs: php4 /ext/standard/tests/file userstreams.phpt
At 16:36 29.10.2002, Wez Furlong wrote:
On 10/29/02, Marcus Börger [EMAIL PROTECTED] wrote:
This test fails with adding
--INI--
output_buffering=0
output_handler=
zlib.output_compression=
implicit_flush=0
Works fine here.
and even when removing the three calls to ob_xxx89
Can you email me the output?
There is no output
Also, could you strace it too? (just to be sure)
[marcus@zaphod php4-HEAD]$ strace php ext/standard/tests/file/userstreams.php
execve(/usr/src/php4-HEAD/sapi/cli/php, [php,
ext/standard/tests/file/userstreams.php], [/* 39 vars */]) = 0
uname({sys=Linux, node=zaphod, ...}) = 0
brk(0) = 0x82886e4
open(/etc/ld.so.preload, O_RDONLY)= -1 ENOENT (No such file or directory)
open(/usr/local/lib/i686/mmx/libsablot.so.0, O_RDONLY) = -1 ENOENT (No
such file or directory)
stat64(/usr/local/lib/i686/mmx, 0xbfffecdc) = -1 ENOENT (No such file or
directory)
open(/usr/local/lib/i686/libsablot.so.0, O_RDONLY) = -1 ENOENT (No such
file or directory)
stat64(/usr/local/lib/i686, 0xbfffecdc) = -1 ENOENT (No such file or
directory)
open(/usr/local/lib/mmx/libsablot.so.0, O_RDONLY) = -1 ENOENT (No such
file or directory)
stat64(/usr/local/lib/mmx, 0xbfffecdc) = -1 ENOENT (No such file or
directory)
open(/usr/local/lib/libsablot.so.0, O_RDONLY) = 3
read(3, \177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0\20H\4\000...,
1024) = 1024
fstat64(3, {st_mode=S_IFREG|0755, st_size=2611187, ...}) = 0
old_mmap(NULL, 1062104, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x40016000
mprotect(0x400eb000, 189656, PROT_NONE) = 0
old_mmap(0x400eb000, 188416, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED,
3, 0xd4000) = 0x400eb000
old_mmap(0x40119000, 1240, PROT_READ|PROT_WRITE,
MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x40119000
close(3)= 0
open(/usr/local/lib/libjs.so, O_RDONLY) = -1 ENOENT (No such file or
directory)
open(/usr/X11R6/lib/i686/mmx/libjs.so, O_RDONLY) = -1 ENOENT (No such
file or directory)
stat64(/usr/X11R6/lib/i686/mmx, 0xbfffeccc) = -1 ENOENT (No such file or
directory)
open(/usr/X11R6/lib/i686/libjs.so, O_RDONLY) = -1 ENOENT (No such file or
directory)
stat64(/usr/X11R6/lib/i686, 0xbfffeccc) = -1 ENOENT (No such file or
directory)
open(/usr/X11R6/lib/mmx/libjs.so, O_RDONLY) = -1 ENOENT (No such file or
directory)
stat64(/usr/X11R6/lib/mmx, 0xbfffeccc) = -1 ENOENT (No such file or
directory)
open(/usr/X11R6/lib/libjs.so, O_RDONLY) = -1 ENOENT (No such file or
directory)
stat64(/usr/X11R6/lib, {st_mode=S_IFDIR|0755, st_size=4096, ...}) = 0
open(/etc/ld.so.cache, O_RDONLY) = 3
fstat64(3, {st_mode=S_IFREG|0644, st_size=53720, ...}) = 0
old_mmap(NULL, 53720, PROT_READ, MAP_PRIVATE, 3, 0) = 0x4011a000
close(3)= 0
open(/usr/lib/libjs.so, O_RDONLY) = 3
read(3, \177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0\300u\1..., 1024)
= 1024
fstat64(3, {st_mode=S_IFREG|0755, st_size=445580, ...}) = 0
old_mmap(NULL, 449188, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x40128000
mprotect(0x40192000, 15012, PROT_NONE) = 0
old_mmap(0x40192000, 16384, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 3,
0x69000) = 0x40192000
close(3)= 0
open(/usr/local/lib/libexpat.so.0, O_RDONLY) = -1 ENOENT (No such file or
directory)
open(/usr/X11R6/lib/libexpat.so.0, O_RDONLY) = -1 ENOENT (No such file or
directory)
open(/usr/lib/libexpat.so.0, O_RDONLY) = 3
read(3, \177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0@%\0\000..., 1024)
= 1024
fstat64(3, {st_mode=S_IFREG|0755, st_size=178576, ...}) = 0
old_mmap(NULL, 181656, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x40196000
mprotect(0x401b9000, 38296, PROT_NONE) = 0
old_mmap(0x401b9000, 40960, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 3,
0x22000) = 0x401b9000
close(3)= 0
open(/usr/local/lib/libpq.so.2, O_RDONLY) = -1 ENOENT (No such file or
directory)
open(/usr/X11R6/lib/libpq.so.2, O_RDONLY) = -1 ENOENT (No such file or
directory)
open(/usr/lib/libpq.so.2, O_RDONLY) = 3
read(3, \177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0\3004\0..., 1024)
= 1024
fstat64(3, {st_mode=S_IFREG|0755, st_size=71712, ...}) = 0
old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1,
0) = 0x401c3000
old_mmap(NULL, 74816, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x401c4000
mprotect(0x401d5000, 5184, PROT_NONE) = 0
old_mmap(0x401d5000, 8192, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 3,
0x1) = 0x401d5000
close(3)= 0
open(/usr/local/lib/libmhash.so.2, O_RDONLY) = -1 ENOENT (No such file or
directory)
open(/usr/X11R6/lib/libmhash.so.2, O_RDONLY) = -1 ENOENT (No such file or
directory)
open(/usr/lib/libmhash.so.2, O_RDONLY) = 3
read(3, \177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0\360\32..., 1024)
= 1024
fstat64(3, {st_mode=S_IFREG|0755, st_size=132160, ...}) = 0
old_mmap(NULL, 135240, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x401d7000
Re: [PHP-DEV] Re: [PHP-CVS] cvs: php4 /ext/standard/tests/file userstreams.phpt
OK, so how about a backtrace ? --Wez. On Tue, 29 Oct 2002, Marcus [iso-8859-1] Börger wrote: At 16:36 29.10.2002, Wez Furlong wrote: Also, could you strace it too? (just to be sure) old_mmap(NULL, 266240, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x407d5000 read(3, ...naq gur ebnq orpbzrf zl oevqr..., 8192) = 8192 lseek(3, 0, SEEK_SET) = 0 --- SIGSEGV (Segmentation fault) --- +++ killed by SIGSEGV +++ -- PHP Development Mailing List http://www.php.net/ To unsubscribe, visit: http://www.php.net/unsub.php
Re: [PHP-DEV] Re: [PHP-CVS] cvs: php4 /ext/standard/tests/file userstreams.phpt
(gdb) file /usr/src/php4-HEAD/sapi/cli/php Reading symbols from /usr/src/php4-HEAD/sapi/cli/php...done. (gdb) set args ext/standard/tests/file/userstreams.phpt (gdb) run Starting program: /usr/src/php4-HEAD/sapi/cli/php ext/standard/tests/file/userstreams.phpt [New Thread 1024 (LWP 13760)] Program received signal SIGSEGV, Segmentation fault. [Switching to Thread 1024 (LWP 13760)] 0x081cb069 in zend_assign_add_obj_handler (execute_data=0xbfffa520, op_array=0x407581bc, tsrm_ls=0x82887b0) at /usr/src/php4-HEAD/Zend/zend_execute.c:232 232 if ((*object_ptr)-type == IS_NULL (gdb) bt #0 0x081cb069 in zend_assign_add_obj_handler (execute_data=0xbfffa520, op_array=0x407581bc, tsrm_ls=0x82887b0) at /usr/src/php4-HEAD/Zend/zend_execute.c:232 #1 0x081c34de in execute (op_array=0x407581bc, tsrm_ls=0x82887b0) at /usr/src/php4-HEAD/Zend/zend_execute.c:1194 #2 0x081a9f7c in call_user_function_ex (function_table=0x4073e74c, object_pp=0x40798984, function_name=0xbfffa6a0, retval_ptr_ptr=0xbfffa694, param_count=1, params=0xbfffa698, no_separation=0, symbol_table=0x0, tsrm_ls=0x82887b0) at /usr/src/php4-HEAD/Zend/zend_execute_API.c:646 #3 0x081944f7 in php_userstreamop_read (stream=0x40799760, buf=0x407cea4c , count=8192, tsrm_ls=0x82887b0) at /usr/src/php4-HEAD/main/user_streams.c:444 #4 0x0818cdb9 in php_stream_fill_read_buffer (stream=0x40799760, size=1023, tsrm_ls=0x82887b0) at /usr/src/php4-HEAD/main/streams.c:491 #5 0x0818d1e0 in _php_stream_get_line (stream=0x40799760, buf=0x407ce610 , maxlen=1024, returned_len=0xbfffa770, tsrm_ls=0x82887b0) at /usr/src/php4-HEAD/main/streams.c:758 #6 0x08120a1c in zif_fgets (ht=2, return_value=0x407cc368, this_ptr=0x0, return_value_used=1, tsrm_ls=0x82887b0) at /usr/src/php4-HEAD/ext/standard/file.c:1341 #7 0x081d4bfa in zend_do_fcall_common_helper (execute_data=0xbfffd7b0, op_array=0x40738864, tsrm_ls=0x82887b0) at /usr/src/php4-HEAD/Zend/zend_execute.c:2414 ---Type return to continue, or q return to quit--- #8 0x081d5419 in zend_do_fcall_handler (execute_data=0xbfffd7b0, op_array=0x40738864, tsrm_ls=0x82887b0) at /usr/src/php4-HEAD/Zend/zend_execute.c:2533 #9 0x081c34de in execute (op_array=0x40738864, tsrm_ls=0x82887b0) at /usr/src/php4-HEAD/Zend/zend_execute.c:1194 #10 0x081b3177 in zend_execute_scripts (type=8, tsrm_ls=0x82887b0, retval=0x0, file_count=3) at /usr/src/php4-HEAD/Zend/zend.c:862 #11 0x08183f99 in php_execute_script (primary_file=0xbab0, tsrm_ls=0x82887b0) at /usr/src/php4-HEAD/main/main.c:1541 #12 0x081dfa30 in main (argc=2, argv=0xbb44) at /usr/src/php4-HEAD/sapi/cli/php_cli.c:695 #13 0x40561280 in __libc_start_main () from /lib/libc.so.6 (gdb) At 19:02 29.10.2002, you wrote: OK, so how about a backtrace ? --Wez. On Tue, 29 Oct 2002, Marcus [iso-8859-1] Börger wrote: At 16:36 29.10.2002, Wez Furlong wrote: Also, could you strace it too? (just to be sure) old_mmap(NULL, 266240, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x407d5000 read(3, ...naq gur ebnq orpbzrf zl oevqr..., 8192) = 8192 lseek(3, 0, SEEK_SET) = 0 --- SIGSEGV (Segmentation fault) --- +++ killed by SIGSEGV +++ -- PHP Development Mailing List http://www.php.net/ To unsubscribe, visit: http://www.php.net/unsub.php
Re: [PHP-DEV] Re: [PHP-CVS] cvs: php4 /ext/standard/tests/file userstreams.phpt
Is that ZE2? There is no zend_assign_add_obj_handler function anywhere in my checkout of HEAD + ZE1, and line 232 of zend_execute is a case statement in my checkout. I'm not concerned with ZE2 problems at this time (well, I am, but don't have the time to support it). Most likely, this is an issue with the recent splitting of the opcode handlers; Andi would probably like to see this backtrace. --Wez. On 29/10/02, Marcus Börger [EMAIL PROTECTED] wrote: (gdb) file /usr/src/php4-HEAD/sapi/cli/php Reading symbols from /usr/src/php4-HEAD/sapi/cli/php...done. (gdb) set args ext/standard/tests/file/userstreams.phpt (gdb) run Starting program: /usr/src/php4-HEAD/sapi/cli/php ext/standard/tests/file/userstreams.phpt [New Thread 1024 (LWP 13760)] Program received signal SIGSEGV, Segmentation fault. [Switching to Thread 1024 (LWP 13760)] 0x081cb069 in zend_assign_add_obj_handler (execute_data=0xbfffa520, op_array=0x407581bc, tsrm_ls=0x82887b0) at /usr/src/php4-HEAD/Zend/zend_execute.c:232 232 if ((*object_ptr)-type == IS_NULL (gdb) bt #0 0x081cb069 in zend_assign_add_obj_handler (execute_data=0xbfffa520, op_array=0x407581bc, tsrm_ls=0x82887b0) at /usr/src/php4-HEAD/Zend/zend_execute.c:232 #1 0x081c34de in execute (op_array=0x407581bc, tsrm_ls=0x82887b0) at /usr/src/php4-HEAD/Zend/zend_execute.c:1194 #2 0x081a9f7c in call_user_function_ex (function_table=0x4073e74c, object_pp=0x40798984, function_name=0xbfffa6a0, retval_ptr_ptr=0xbfffa694, param_count=1, params=0xbfffa698, no_separation=0, symbol_table=0x0, tsrm_ls=0x82887b0) at /usr/src/php4-HEAD/Zend/zend_execute_API.c:646 #3 0x081944f7 in php_userstreamop_read (stream=0x40799760, buf=0x407cea4c , count=8192, tsrm_ls=0x82887b0) at /usr/src/php4-HEAD/main/user_streams.c:444 #4 0x0818cdb9 in php_stream_fill_read_buffer (stream=0x40799760, size=1023, tsrm_ls=0x82887b0) at /usr/src/php4-HEAD/main/streams.c:491 #5 0x0818d1e0 in _php_stream_get_line (stream=0x40799760, buf=0x407ce610 , maxlen=1024, returned_len=0xbfffa770, tsrm_ls=0x82887b0) at /usr/src/php4-HEAD/main/streams.c:758 #6 0x08120a1c in zif_fgets (ht=2, return_value=0x407cc368, this_ptr=0x0, return_value_used=1, tsrm_ls=0x82887b0) at /usr/src/php4-HEAD/ext/standard/file.c:1341 #7 0x081d4bfa in zend_do_fcall_common_helper (execute_data=0xbfffd7b0, op_array=0x40738864, tsrm_ls=0x82887b0) at /usr/src/php4-HEAD/Zend/zend_execute.c:2414 ---Type return to continue, or q return to quit--- #8 0x081d5419 in zend_do_fcall_handler (execute_data=0xbfffd7b0, op_array=0x40738864, tsrm_ls=0x82887b0) at /usr/src/php4-HEAD/Zend/zend_execute.c:2533 #9 0x081c34de in execute (op_array=0x40738864, tsrm_ls=0x82887b0) at /usr/src/php4-HEAD/Zend/zend_execute.c:1194 #10 0x081b3177 in zend_execute_scripts (type=8, tsrm_ls=0x82887b0, retval=0x0, file_count=3) at /usr/src/php4-HEAD/Zend/zend.c:862 #11 0x08183f99 in php_execute_script (primary_file=0xbab0, tsrm_ls=0x82887b0) at /usr/src/php4-HEAD/main/main.c:1541 #12 0x081dfa30 in main (argc=2, argv=0xbb44) at /usr/src/php4-HEAD/sapi/cli/php_cli.c:695 #13 0x40561280 in __libc_start_main () from /lib/libc.so.6 (gdb) At 19:02 29.10.2002, you wrote: OK, so how about a backtrace ? --Wez. On Tue, 29 Oct 2002, Marcus [iso-8859-1] Börger wrote: At 16:36 29.10.2002, Wez Furlong wrote: Also, could you strace it too? (just to be sure) old_mmap(NULL, 266240, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x407d5000 read(3, ...naq gur ebnq orpbzrf zl oevqr..., 8192) = 8192 lseek(3, 0, SEEK_SET) = 0 --- SIGSEGV (Segmentation fault) --- +++ killed by SIGSEGV +++ -- PHP Development Mailing List http://www.php.net/ To unsubscribe, visit: http://www.php.net/unsub.php
