t/src/extend.t: failing in trunk

[jkeenan]

At some point in the past 24 hours (r41766 or earlier), this file began to experience failures in trunk:

not ok 15 - compile string in a fresh interp - \#39986

#   Failed test 'compile string in a fresh interp - \#39986'
#   at t/src/extend.t line 624.
# Exited with error code: [SIGNAL 134]
# Received:
# src/string/api.c:549: failed assertion '(b)->encoding'
# 
# Expected:
# Hello from foo!
# 
...
# Looks like you failed 1 test of 17.

According to Smolder reports, this is happening on some, but not all, OSes.

Is there someone awake who could bisect/diagnose?

Thank you very much.

kid51

[jkeenan]

Replying to jkeenan:

At some point in the past 24 hours (r41766 or earlier), this file began to experience failures in trunk:

The attached change in src/embed.c -- part of a much larger commit -- may be a good place to look for the source of this problem.

kid51

[jkeenan]

svn diff -r 41764 src/embed.c

[bacek]

Replying to jkeenan:

The attached change in src/embed.c -- part of a much larger commit -- may be a good place to look for the source of this problem.

It was my commit. But I'm pretty sure that it didn't break it so badly. I'll try to find someone on IRC for interactive debugging to understand what happend.

-- Bacek

[mikehh]

just to confirm the test passes at r41764 and fails at r41765

[bacek]

Test fixed in r41776. Closing ticket.

[jkeenan]

It's always helpful to have the OP confirm:

$ prove -v t/src/extend.t
t/src/extend.t .. 
1..17
ok 1 - set/get_intreg
...
ok 15 - compile string in a fresh interp - \#39986
not ok 16 - call multi sub from C - \#41511 # TODO RT #41511

#   Failed (TODO) test 'call multi sub from C - \#41511'
#   at t/src/extend.t line 666.
# Exited with error code: 1
# Received:
# Attempting to do sub operation on non-Sub.
# current instr.: 'main' pc 0 (/tmp/yWVl3uNz6f.pir:1)
# 
# Expected:
# Result is 300.
# 
ok 17 - multiple Parrot_new/Parrot_exit cycles
ok
All tests successful.
Files=1, Tests=17, 25 wallclock secs ( 0.12 usr  0.05 sys +  4.76 cusr  4.28 csys =  9.21 CPU)
Result: PASS

[bacek]

Replying to jkeenan:

It's always helpful to have the OP confirm:

Oh which architecture is it? I checked with darbelo++ on IRC for OpenBSD and linux/amd64 report from smolder before closing ticket.

-- Bacek

[jkeenan]

Replying to bacek:

Replying to jkeenan:

It's always helpful to have the OP confirm:

Oh which architecture is it? I checked with darbelo++ on IRC for OpenBSD and linux/amd64 report from smolder before closing ticket.

Darwin/PPC. There was no need to re-open the ticket. Test 15, the subject of the original post, once again passes. (Test 16 remains TODO-ed.) Re-closing.