You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@httpd.apache.org by Stefan Eissing <st...@greenbytes.de> on 2017/01/30 10:04:57 UTC
Re: Strange http2 test error on Solaris with static linking
Rainer,
this was sitting on my "to be investigated" pile for some time now. Sorry about that.
The log line indicates a code restructuring in processing responses that has since gotten 2 fixes in regard exactly such log lines. If you by occasion re-run the recent trunk/2.4.x again, I'd like to hear back if this is now also resolved. Thanks!
Cheers, Stefan
> Am 19.12.2016 um 23:08 schrieb Rainer Jung <ra...@kippdata.de>:
>
> The test t/modules/http2.t fails consistently on Solaris when httpd is statically linked. Results are based on 2.4.25.
>
> Here is an excerpt from the truss output for the first test in http2.t (all fail) starting around the first difference between the good and bad case. The first log line indicating the problem is:
>
> [Mon Dec 19 22:39:20.178603 2016] [http2:debug] [pid 9984:tid 52] h2_session.c(1592): [client 127.0.0.1:64013] AH03466: h2_stream(215-1): no response, RST_STREAM, err=1
>
> Fail:
>
> 9984/20: open(".../t/htdocs/index.html", O_RDONLY|O_LARGEFILE) = 39
> 9984/20: fcntl(39, F_GETFD, 0x00000000) = 0
> 9984/20: fcntl(39, F_SETFD, 0x00000001) = 0
> 9984/20: getpid() = 9984 [9960]
> 9984/20: write(2, 0xFCE77970, 134) = 134
> 9984/20: [ M o n D e c 1 9 2 2 : 3 9 : 2 0 . 1 7 7 4 8 8 2 0 1 6
> 9984/20: ] [ h e a d e r s : t r a c e 2 ] [ p i d 9 9 8 4 : t i d
> 9984/20: 2 0 ] m o d _ h e a d e r s . c ( 8 7 5 ) : A H 0 1 5 0 2
> 9984/20: : h e a d e r s : a p _ h e a d e r s _ o u t p u t _ f i l
> 9984/20: t e r ( )\n
> 9984/20: getpid() = 9984 [9960]
> 9984/20: write(2, 0xFCE76AA8, 167) = 167
> 9984/20: [ M o n D e c 1 9 2 2 : 3 9 : 2 0 . 1 7 7 9 0 9 2 0 1 6
> 9984/20: ] [ h t t p 2 : d e b u g ] [ p i d 9 9 8 4 : t i d 2 0
> 9984/20: ] h 2 _ t a s k . c ( 8 8 ) : [ c l i e n t 1 2 7 . 0 . 0
> 9984/20: . 1 : 6 4 0 1 3 ] A H 0 3 3 4 8 : h 2 _ t a s k ( 2 1 5 - 1
> 9984/20: ) : o p e n o u t p u t t o G E T l o c a l h o s t :
> 9984/20: 8 5 4 6 /\n
> 9984/20: lwp_unpark(52) = 0
> 9984/52: lwp_park(0xFAE7BAA8, 0) Err#62 ETIME
> 9984/52: timeout: 0.000000000 sec
> 9984/20: lwp_unpark(52) = 0
> 9984/52: lwp_park(0x00000000, 0) = 0
> 9984/52: read(38, 0x00684688, 8000) Err#11 EAGAIN
> 9984/52: getpid() = 9984 [9960]
> 9984/52: write(2, 0xFAE77938, 169) = 169
> 9984/52: [ M o n D e c 1 9 2 2 : 3 9 : 2 0 . 1 7 8 6 0 3 2 0 1 6
> 9984/52: ] [ h t t p 2 : d e b u g ] [ p i d 9 9 8 4 : t i d 5 2
> 9984/52: ] h 2 _ s e s s i o n . c ( 1 5 9 2 ) : [ c l i e n t 1 2
> 9984/52: 7 . 0 . 0 . 1 : 6 4 0 1 3 ] A H 0 3 4 6 6 : h 2 _ s t r e a
> 9984/52: m ( 2 1 5 - 1 ) : n o r e s p o n s e , R S T _ S T R E A
> 9984/52: M , e r r = 1\n
> 9984/52: getpid() = 9984 [9960]
> 9984/52: write(2, 0xFAE778D0, 206) = 206
> 9984/52: [ M o n D e c 1 9 2 2 : 3 9 : 2 0 . 1 7 8 8 2 5 2 0 1 6
> 9984/52: ] [ h t t p 2 : d e b u g ] [ p i d 9 9 8 4 : t i d 5 2
> 9984/52: ] h 2 _ s e s s i o n . c ( 6 6 1 ) : [ c l i e n t 1 2 7
> 9984/52: . 0 . 0 . 1 : 6 4 0 1 3 ] A H 0 3 0 6 8 : h 2 _ s e s s i o
> 9984/52: n ( 2 1 5 ) : s e n t F R A M E [ R S T _ S T R E A M [ l e
> 9984/52: n g t h = 4 , f l a g s = 0 , s t r e a m = 1 ] ] , f r a
> 9984/52: m e s = 3 / 3 ( r / s )\n
>
>
> OK:
>
> 10047/11: open(".../t/htdocs/index.html", O_RDONLY|O_LARGEFILE) = 39
> 10047/11: fcntl(39, F_GETFD, 0x00000000) = 0
> 10047/11: fcntl(39, F_SETFD, 0x00000001) = 0
> 10047/11: getpid() = 10047 [10039]
> 10047/11: write(2, 0xFC377970, 135) = 135
> 10047/11: [ M o n D e c 1 9 2 2 : 4 1 : 3 6 . 9 1 6 8 5 9 2 0 1 6
> 10047/11: ] [ h e a d e r s : t r a c e 2 ] [ p i d 1 0 0 4 7 : t i
> 10047/11: d 1 1 ] m o d _ h e a d e r s . c ( 8 7 5 ) : A H 0 1 5 0
> 10047/11: 2 : h e a d e r s : a p _ h e a d e r s _ o u t p u t _ f i
> 10047/11: l t e r ( )\n
> 10047/52: lwp_park(0xF9A7BAA8, 0) Err#62 ETIME
> 10047/52: timeout: 0.000000000 sec
> 10047/52: read(38, 0x003196D8, 8000) Err#11 EAGAIN
> 10047/11: getpid() = 10047 [10039]
> 10047/52: getpid() = 10047 [10039]
> 10047/11: write(2, 0xFC376AA8, 167) = 167
> 10047/11: [ M o n D e c 1 9 2 2 : 4 1 : 3 6 . 9 3 6 5 4 5 2 0 1 6
> 10047/11: ] [ h t t p 2 : d e b u g ] [ p i d 1 0 0 4 7 : t i d 1
> 10047/11: 1 ] h 2 _ t a s k . c ( 8 8 ) : [ c l i e n t 1 2 7 . 0 .
> 10047/11: 0 . 1 : 6 4 0 1 8 ] A H 0 3 3 4 8 : h 2 _ t a s k ( 8 7 - 1
> 10047/11: ) : o p e n o u t p u t t o G E T l o c a l h o s t :
> 10047/11: 8 5 4 6 /\n
> 10047/52: write(2, 0xF9A77A00, 166) = 166
> 10047/52: [ M o n D e c 1 9 2 2 : 4 1 : 3 6 . 9 3 6 5 7 0 2 0 1 6
> 10047/52: ] [ c o r e : t r a c e 6 ] [ p i d 1 0 0 4 7 : t i d 5
> 10047/52: 2 ] c o r e _ f i l t e r s . c ( 5 2 5 ) : [ c l i e n t
> 10047/52: 1 2 7 . 0 . 0 . 1 : 6 4 0 1 8 ] c o r e _ o u t p u t _ f i l
> 10047/52: t e r : f l u s h i n g b e c a u s e o f F L U S H b
> 10047/52: u c k e t\n
> 10047/52: read(38, 0x003196D8, 8000) Err#11 EAGAIN
> 10047/52: getpid() = 10047 [10039]
> 10047/52: write(2, 0xF9A77920, 184) = 184
> 10047/52: [ M o n D e c 1 9 2 2 : 4 1 : 3 6 . 9 3 7 0 9 7 2 0 1 6
> 10047/52: ] [ h t t p 2 : d e b u g ] [ p i d 1 0 0 4 7 : t i d 5
> 10047/52: 2 ] h 2 _ s e s s i o n . c ( 1 4 5 7 ) : [ c l i e n t 1
> 10047/52: 2 7 . 0 . 0 . 1 : 6 4 0 1 8 ] A H 0 3 0 7 3 : h 2 _ s t r e
> 10047/52: a m ( 8 7 - 1 ) : s u b m i t r e s p o n s e 2 0 0 , R
> 10047/52: E M O T E _ W I N D O W _ S I Z E = 6 5 5 3 5\n
>
> The communication seems to work in principle, so at least no obvious SSL compatibility problem. Any ideas?
>
> Regards,
>
> Rainer
Stefan Eissing
<green/>bytes GmbH
Hafenstrasse 16
48155 Münster
www.greenbytes.de