Bug#644163: [SOLVED] Bug#644163: libv8: FTBFS on armel: 12 tests failed

[Jérémy Lal]

very nice of you,
will upload fixed version asap.



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Bug#644163: [SOLVED] Bug#644163: libv8: FTBFS on armel: 12 tests failed

[Timo Juhani Lindfors]

Hi,

here's a fix that works for me:

Author: Vyacheslav Egorov
Origin: 
https://raw.github.com/gist/1302061/371cebbda9e2bf0790728f3b5f47d38b74a26368/CountLeadingZeros.patch
Debian-Bug: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=644163

Index: src/arm/macro-assembler-arm.cc
===
--- a/src/arm/macro-assembler-arm.cc(revision 9710)
+++ b/src/arm/macro-assembler-arm.cc(working copy)
@@ -2907,8 +2907,8 @@
 #ifdef CAN_USE_ARMV5_INSTRUCTIONS
   clz(zeros, source);  // This instruction is only supported after
   ARM5.
 #else
+  Move(scratch, source);
   mov(zeros, Operand(0, RelocInfo::NONE));
-  Move(scratch, source);
   // Top 16.
   tst(scratch, Operand(0x));
   add(zeros, zeros, Operand(16), LeaveCC, eq);


Vyacheslav said on IRC that he'll commit this to repository soon.

-Timo



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Bug#644163: libv8: FTBFS on armel: 12 tests failed

[Timo Juhani Lindfors]

Hi,

I minimized the regress-1122 test failure to

$ shell -e 'function x1() { n = 16; return (n / 2); } print(x1()); print(x1());'
8
16

As suggested on #v8 I rebuilt using objectprint=on disassembler=on and
saved the generated code:

$ shell --print-code -e 'function x1() { n = 16; return (n / 2); } print(x1()); 
print(x1());' > print-code.txt
$ shell --print-code-stubs -e 'function x1() { n = 16; return (n / 2); } 
print(x1()); print(x1());' > print-code-stubs.txt

Both files are attached to this email.

--- Raw source ---
function x1() { n = 16; return (n / 2); } print(x1()); print(x1());

--- Code ---
kind = FUNCTION
Instructions (size = 260)
0x577fa160 0  e92d4902   stmdb sp!, {r1, r8, fp, lr}
0x577fa164 4  e59ac00c   ldr ip, [r10, #+12]
0x577fa168 8  e28db008   add fp, sp, #8
0x577fa16c12  e52dc004   str ip, [sp, #-4]!
0x577fa17016  e59f20b8   ldr r2, [pc, #+184] ;; object: 
0x567b73d9 
0x577fa17420  e3a01000   mov r1, #0
0x577fa17824  e3a0   mov r0, #0
0x577fa17c28  e92d0107   stmdb sp!, {r0, r1, r2, r8}
0x577fa18032  e3a4   mov r0, #4
0x577fa18436  e59f10a8   ldr r1, [pc, #+168]
0x577fa18840  e1a0e00f   mov lr, pc
0x577fa18c44  e59ff0a4   ldr pc, [pc, #+164] ;; debug: 
statement 0
 ;; code: STUB, 
CEntryStub, minor: 0
0x577fa19048  e59ac044   ldr ip, [r10, #+68]
0x577fa19452  e15d000c   cmp sp, ip
0x577fa19856  2a01   bcs +12 -> 68  (0x577fa1a4)
0x577fa19c60  e1a0e00f   mov lr, pc
0x577fa1a064  e59ff094   ldr pc, [pc, #+148] ;; code: STUB, 
StackCheckStub, minor: 0
0x577fa1a468  e5980013   ldr r0, [r8, #+19]
0x577fa1a872  e52d0004   str r0, [sp, #-4]!
0x577fa1ac76  e5980013   ldr r0, [r8, #+19]
0x577fa1b080  e52d0004   str r0, [sp, #-4]!
0x577fa1b484  e59f2084   ldr r2, [pc, #+132] ;; object: 
0x2cd26e59 
0x577fa1b888  e1a0e00f   mov lr, pc
0x577fa1bc92  e59ff080   ldr pc, [pc, #+128] ;; debug: 
statement 42
 ;; debug: position 
48
 ;; code: 
contextual, CALL_IC, UNINITIALIZED, argc = 0
0x577fa1c096  e51b8004   ldr r8, [fp, #-4]
0x577fa1c4   100  e52d0004   str r0, [sp, #-4]!
0x577fa1c8   104  e59f2078   ldr r2, [pc, #+120] ;; object: 
0x2cd26a41 
0x577fa1cc   108  e1a0e00f   mov lr, pc
0x577fa1d0   112  e59ff074   ldr pc, [pc, #+116] ;; debug: 
statement 42
 ;; code: 
contextual, CALL_IC, UNINITIALIZED, argc = 1
0x577fa1d4   116  e51b8004   ldr r8, [fp, #-4]
0x577fa1d8   120  e5980013   ldr r0, [r8, #+19]
0x577fa1dc   124  e52d0004   str r0, [sp, #-4]!
0x577fa1e0   128  e5980013   ldr r0, [r8, #+19]
0x577fa1e4   132  e52d0004   str r0, [sp, #-4]!
0x577fa1e8   136  e59f2060   ldr r2, [pc, #+96]  ;; object: 
0x2cd26e59 
0x577fa1ec   140  e1a0e00f   mov lr, pc
0x577fa1f0   144  e59ff05c   ldr pc, [pc, #+92]  ;; debug: 
statement 55
 ;; debug: position 
61
 ;; code: 
contextual, CALL_IC, UNINITIALIZED, argc = 0
0x577fa1f4   148  e51b8004   ldr r8, [fp, #-4]
0x577fa1f8   152  e52d0004   str r0, [sp, #-4]!
0x577fa1fc   156  e59f2054   ldr r2, [pc, #+84]  ;; object: 
0x2cd26a41 
0x577fa200   160  e1a0e00f   mov lr, pc
0x577fa204   164  e59ff050   ldr pc, [pc, #+80]  ;; debug: 
statement 55
 ;; code: 
contextual, CALL_IC, UNINITIALIZED, argc = 1
0x577fa208   168  e51b8004   ldr r8, [fp, #-4]
0x577fa20c   172  e50b000c   str r0, [fp, #-12]
0x577fa210   176  e51b000c   ldr r0, [fp, #-12]
0x577fa214   180  e1a0d00b   mov sp, fp  ;; debug: 
statement 66
 ;; js return
0x577fa218   184  e8bd4800   ldmia sp!, {fp, lr}
0x577fa21c   188  e28dd004   add sp, sp, #4
0x577fa220   192  e1a0f00e   mov pc, lr
0x577fa224   196  e59a000c   ldr r0, [r10, #+12]
0x577fa228   200  eaf9   b -20 -> 180  (0x577fa214)
0x577fa22c   204  0c0c   constant pool begin
0x577fa230   208  567b73d9   constant
0x577fa234   212  40206758   constant
0x577fa238   216  3d688040   constant
0x577fa23c   220  2246f860   constant
0x577fa240   224  2cd26e59   constant
0x577fa244   228  22472120   constant
0x577fa248   232  2cd26a41   constant
0x577fa24c   236  22472040   constant
0x577fa250   240  2cd26e59   constant
0x577fa254   244  22472120   constant
0x577fa258   248  2cd26a41   constant
0x577fa25c   252  22

Bug#644163: libv8: FTBFS on armel: 12 tests failed

[Jérémy Lal]

On 18/10/2011 11:20, Hector Oron wrote:
> Hello,
> 
> On Mon, Oct 03, 2011 at 04:03:37PM +0200, Jérémy Lal wrote:
>> On 03/10/2011 15:48, Julian Andres Klode wrote:
>>> Version: 3.4.14.21-3 
>>> libv8 currently fails to build from source on armel due to failing tests.
>>
>> Hi,
>> i'm well aware of the situation.
>> It has improved a bit : armel 3.4.14.21-3 actually /run/ tests.
> 
> Indeed, a rebuild has happened, just in case you want to cross check with 
> previous build:
>   
> 

I receive those reports, but thank you for your support.
 
>> I have no clue yet as why those tests fail. Things to be verified :
>> * run the tests using d8, not shell, as it is deprecated
>> * build v8 using gyp instead of scons, deprecated too (i am working on it)
>> * update to latest 3.4 revision
>> * check other options : crankshaft, snapshot. They might need to be
>>   toggled to allow armel support.
>>
>> In any case, i'd be more productive with an actual armel shell.
> 
> May I suggest you to apply for a Debian guest account on abel.debian.org:
>   


I sent that (signed) request on 29/08/2011 00:32, i guess it has been ignored :

> Hi,
> here's my request to get an access to armel porter machine(s) :
> 
> First/last name   Jérémy Lal
> Preferred usernamekapouer (i may already have kapouer-guest registered 
> for anonscm.d.o access)
> GPG fingerprint   074141DC
> email address kapo...@melix.org
> DMUP signed agreement 
> http://lists.debian.org/debian-newmaint/2010/04/msg00018.html
> Machines/Architecturesabel, agricola for armel arch
> Rationale libv8 and nodejs FTBFS often with unreproducible bugs 
> under qemu.
>   having access to real armel hardware would help a lot 
> (i'm the
>   leading maintainer of these packages).
> 
> Regards,
> Jérémy.



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Bug#644163: libv8: FTBFS on armel: 12 tests failed

[Hector Oron]

Hello,

On Mon, Oct 03, 2011 at 04:03:37PM +0200, Jérémy Lal wrote:
> On 03/10/2011 15:48, Julian Andres Klode wrote:
> > Version: 3.4.14.21-3 
> > libv8 currently fails to build from source on armel due to failing tests.
> 
> Hi,
> i'm well aware of the situation.
> It has improved a bit : armel 3.4.14.21-3 actually /run/ tests.

Indeed, a rebuild has happened, just in case you want to cross check with 
previous build:
  


> I have no clue yet as why those tests fail. Things to be verified :
> * run the tests using d8, not shell, as it is deprecated
> * build v8 using gyp instead of scons, deprecated too (i am working on it)
> * update to latest 3.4 revision
> * check other options : crankshaft, snapshot. They might need to be
>   toggled to allow armel support.
> 
> In any case, i'd be more productive with an actual armel shell.

May I suggest you to apply for a Debian guest account on abel.debian.org:
  

Best regards,
-- 
 Héctor Orón

"Our Sun unleashes tremendous flares expelling hot gas into the Solar System, 
which one day will disconnect us."

-- Day DVB-T stop working nicely
Video flare: http://antwrp.gsfc.nasa.gov/apod/ap100510.html


signature.asc
Description: Digital signature

Bug#644163: libv8: FTBFS on armel: 12 tests failed

[Jérémy Lal]

On 03/10/2011 15:48, Julian Andres Klode wrote:
> Version: 3.4.14.21-3 
> libv8 currently fails to build from source on armel due to failing tests.

Hi,
i'm well aware of the situation.
It has improved a bit : armel 3.4.14.21-3 actually /run/ tests.

I have no clue yet as why those tests fail. Things to be verified :
* run the tests using d8, not shell, as it is deprecated
* build v8 using gyp instead of scons, deprecated too (i am working on it)
* update to latest 3.4 revision
* check other options : crankshaft, snapshot. They might need to be
  toggled to allow armel support.

In any case, i'd be more productive with an actual armel shell.

Cheers,
Jérémy.



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Bug#644163: libv8: FTBFS on armel: 12 tests failed

[Julian Andres Klode]

Package: libv8
Version: 3.4.14.21-3
Severity: important

libv8 currently fails to build from source on armel due to 
failing tests. The complete build log can be found at

  
https://buildd.debian.org/status/fetch.php?pkg=libv8&arch=armel&ver=3.4.14.21-3&stamp=1317252599

It would be great if that could be fixed, as I want
to install chromium some day, and that waits for
libv8.

-- System Information:
Debian Release: wheezy/sid
  APT prefers unstable
  APT policy: (550, 'unstable'), (550, 'stable'), (500, 'testing'), (300, 
'experimental')
Architecture: amd64 (x86_64)

Kernel: Linux 3.0.0-1-amd64 (SMP w/4 CPU cores)
Locale: LANG=en_GB.utf8, LC_CTYPE=en_GB.utf8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash

-- 
Julian Andres Klode  - Debian Developer, Ubuntu Member

See http://wiki.debian.org/JulianAndresKlode and http://jak-linux.org/.


pgppMAM55NPLC.pgp
Description: PGP signature