www.digitalmars.com         C & C++   DMDScript  

digitalmars.D.bugs - DStress report for DMD-0.127 -> DMD-0.128

reply Thomas Kuehne <thomas-dloop kuehne.this-is-spam.cn> writes:
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

DMD 0.127 -> DMD 0.128 (Linux)

FAIL -> ERROR
http://dstress.kuehne.cn/run/with_13.d
http://dstress.kuehne.cn/nocompile/o/opAssign_02_D.d

XFAIL -> ERROR
http://dstress.kuehne.cn/nocompile/__TIME__02.d
http://dstress.kuehne.cn/nocompile/__TIMESTAMP__02.d
http://dstress.kuehne.cn/nocompile/o/opAssign_01_A.d
http://dstress.kuehne.cn/nocompile/o/opAssign_01_B.d
http://dstress.kuehne.cn/nocompile/o/opAssign_02_A.d
http://dstress.kuehne.cn/nocompile/o/opAssign_02_B.d
http://dstress.kuehne.cn/nocompile/o/opAssign_03_A.d
http://dstress.kuehne.cn/nocompile/o/cast_28.d
http://dstress.kuehne.cn/nocompile/length_03.d
http://dstress.kuehne.cn/nocompile/f/function_03.d
http://dstress.kuehne.cn/nocompile/__FILE__02.d
http://dstress.kuehne.cn/nocompile/d/delegate_15.d
http://dstress.kuehne.cn/nocompile/__DATE__02.d
http://dstress.kuehne.cn/nocompile/cast_10.d
http://dstress.kuehne.cn/nocompile/associative_array_01.d
http://dstress.kuehne.cn/nocompile/alias_16.d
http://dstress.kuehne.cn/nocompile/alias_15.d

Pass: 1396 (+33)
XPass:  70 (+ 0)
Fail:  187 (-14)
XFail: 602 (-12)
Error:  33 (- 4)

hot spots: (217K)
http://dstress.kuehne.cn/www/dstress.html

complete report: (1,5M)
http://dstress.kuehne.cn/www/results.html

Thomas





-----BEGIN PGP SIGNATURE-----

iD8DBQFC3+/G3w+/yD4P9tIRApaDAJ4pYsesbX9SIJr95jnwdFxT2xb+yACeJHC4
qwcXeD//IkLsru6XiURoEOg=
=OiOS
-----END PGP SIGNATURE-----
Jul 21 2005
parent reply Stewart Gordon <smjg_1998 yahoo.com> writes:
Success rate up from 86.7% to 87.5%
Stability rate down from 98.4% to 98.3%

http://smjg.port5.com/pr/d/dstress.xls

Are the figures for versions that have disappeared off the table still 
available anywhere?  OK, so you may not have retested them recently, but 
the mere fact that some testcases have been added/changed/reclassified 
will have made a difference.

And how are undefined testcases supposed to be listed?  I see 
abstract_01 to 04 have just disappeared, but array_initialization_02 and 
array_initialization_17_A to C are still filed under run.

They certainly should be displayed - highlighting 
ambiguities/inconsistencies in the spec is an essential part of helping 
to get the language up to scratch.

Stewart.

-- 
-----BEGIN GEEK CODE BLOCK-----
Version: 3.1
GCS/M d- s:- a->--- UB  P+ L E  W++  N+++ o K- w++  O? M V? PS- PE- Y? 
PGP- t- 5? X? R b DI? D G e++>++++ h-- r-- !y
------END GEEK CODE BLOCK------

My e-mail is valid but not my primary mailbox.  Please keep replies on 
the 'group where everyone may benefit.
Jul 22 2005
parent reply =?ISO-8859-1?Q?Thomas_K=FChne?= <thomas-dloop kuehne.THISISSPAM.cn> writes:
Stewart Gordon schrieb:

 Success rate up from 86.7% to 87.5%
 Stability rate down from 98.4% to 98.3%

 http://smjg.port5.com/pr/d/dstress.xls

 Are the figures for versions that have disappeared off the table still
 available anywhere?  OK, so you may not have retested them recently, but
 the mere fact that some testcases have been added/changed/reclassified
 will have made a difference.


The raw data is still available:
http://dstress.kuehne.cn/raw_results

How to create a result table:

1) download the test suite
svn co http://dstress.kuehne.cn

2) compile (www/GenReport.java)
javac GenReport.java

3) generate result table (raw_results)
java cn.kuehne.dmd.dstress.GenReport linux-i686_dmd-0.093
linux-i686_dmd-0.095 ... > big_table.html


 And how are undefined testcases supposed to be listed?  I see
 abstract_01 to 04 have just disappeared, but array_initialization_02 and
 array_initialization_17_A to C are still filed under run.


Added link: http://dstress.kuehne.cn/undefined/

The documentation for the array_initialization test cases is logical
consistent, but definitely requires a cleaner rewording.

Thomas
Jul 22 2005
parent reply Stewart Gordon <smjg_1998 yahoo.com> writes:
Thomas Kühne wrote:
<snip>

 The documentation for the array_initialization test cases is logical
 consistent, but definitely requires a cleaner rewording.


I don't understand.  How does

http://www.wikiservice.at/wiki4d/wiki.cgi?DocComments/Arrays#Contradiction

constitute logical consistency?

Stewart.

-- 
-----BEGIN GEEK CODE BLOCK-----
Version: 3.1
GCS/M d- s:- a->--- UB  P+ L E  W++  N+++ o K- w++  O? M V? PS- PE- Y? 
PGP- t- 5? X? R b DI? D G e++>++++ h-- r-- !y
------END GEEK CODE BLOCK------

My e-mail is valid but not my primary mailbox.  Please keep replies on 
the 'group where everyone may benefit.
Jul 25 2005
parent reply Thomas Kühne <Thomas_member pathlink.com> writes:
In article <dc2eit$1hbq$1 digitaldaemon.com>, Stewart Gordon says...

Thomas Kühne wrote:
<snip>

 The documentation for the array_initialization test cases is logical
 consistent, but definitely requires a cleaner rewording.


I don't understand.  How does

http://www.wikiservice.at/wiki4d/wiki.cgi?DocComments/Arrays#Contradiction

constitute logical consistency?


"If any members of an array are initialized."

I'll use an example to explain the context:

type[3] a = [ 1:2, 3 ] => [type.init, 2, 3]; // legal
int[3] a = [ 1:2, 3 ] => [??, 2, 3]; // illegal

It nowhere says that the coder has to initialise all elements. Thus if the coder
doesn't explicitly initialise some array members the compiler has to deal with
them.

The troublesome wording of the reasioning in the doc seems to cause some
confusion.

Thomas
Jul 25 2005
parent Stewart Gordon <smjg_1998 yahoo.com> writes:
Thomas Kühne wrote:

 In article <dc2eit$1hbq$1 digitaldaemon.com>, Stewart Gordon says...

 Thomas Kühne wrote:
 <snip>

 The documentation for the array_initialization test cases is logical
 consistent, but definitely requires a cleaner rewording.


 I don't understand.  How does

 http://www.wikiservice.at/wiki4d/wiki.cgi?DocComments/Arrays#Contradiction

 constitute logical consistency?


 
 "If any members of an array are initialized."
 
 I'll use an example to explain the context:
 
 type[3] a = [ 1:2, 3 ] => [type.init, 2, 3]; // legal
 int[3] a = [ 1:2, 3 ] => [??, 2, 3]; // illegal


Built-in types have default initialisers too.  Moreover, read the sample 
code on that page again.


 It nowhere says that the coder has to initialise all elements. Thus if the
coder
 doesn't explicitly initialise some array members the compiler has to deal with
 them.


<snip>

http://www.digitalmars.com/d/memory.html#uninitializedarrays

"Arrays are always initialized in D."

So if what you say follows, then either

     int[3] a;

would have to be illegal, or the words "If any members of an array are 
initialized" would be redundant.

Stewart.

-- 
-----BEGIN GEEK CODE BLOCK-----
Version: 3.1
GCS/M d- s:- a->--- UB  P+ L E  W++  N+++ o K- w++  O? M V? PS- PE- Y? 
PGP- t- 5? X? R b DI? D G e++>++++ h-- r-- !y
------END GEEK CODE BLOCK------

My e-mail is valid but not my primary mailbox.  Please keep replies on 
the 'group where everyone may benefit.
Jul 25 2005