Re: [openssl.org #3004] Intermittent MASM compilation failures in OpenSSL 1.0.1e VC-WIN64A build

2013-03-18 Thread Andy Polyakov via RT 
> The commit has a typo, the second line should read "*STDOUT=*OUT;".
> 
> The fix works, thanks!

Yes, it was fixed few minutes after. Dismissing the case.


__
OpenSSL Project http://www.openssl.org
Development Mailing List   openssl-dev@openssl.org
Automated List Manager   majord...@openssl.org

Re: Re: [openssl.org #3004] Intermittent MASM compilation failures in OpenSSL 1.0.1e VC-WIN64A build

2013-03-04 Thread jan.evangeli...@seznam.cz via RT 
The commit has a typo, the second line should read "*STDOUT=*OUT;".

The fix works, thanks!

__
OpenSSL Project http://www.openssl.org
Development Mailing List   openssl-dev@openssl.org
Automated List Manager   majord...@openssl.org

Re: [openssl.org #3004] Intermittent MASM compilation failures in OpenSSL 1.0.1e VC-WIN64A build

2013-03-01 Thread Andy Polyakov via RT 
> When building OpenSSL 1.0.1e in the context of a larger build, the build 
> fails with the following error:

See http://rt.openssl.org/Ticket/Display.html?id=2963. Note that there 
were multiple suggestions, but 1st reportedly did the trick.

> perl crypto\bn\asm\x86_64-gf2m.pl tmp32dll\x86_64-gf2m.asm

The fix is in 1.0.1e, but x86_64-gf2m.pl was not fixed, erroneously. 
http://git.openssl.org/gitweb/?p=openssl.git;a=commitdiff;h=bc4ae2cb0b8cf04b64ac91a546439d20926ba569


__
OpenSSL Project http://www.openssl.org
Development Mailing List   openssl-dev@openssl.org
Automated List Manager   majord...@openssl.org

[openssl.org #3004] Intermittent MASM compilation failures in OpenSSL 1.0.1e VC-WIN64A build

2013-03-01 Thread jan.evangeli...@seznam.cz via RT 
When building OpenSSL 1.0.1e in the context of a larger build, the build fails 
with the following error:

perl crypto\bn\asm\x86_64-gf2m.pl tmp32dll\x86_64-gf2m.asm
ml64 /c /Cp /Cx /Zi /Fotmp32dll\x86_64-gf2m.obj tmp32dll\x86_64-gf2m.asm
 Assembling: tmp32dll\x86_64-gf2m.asm
tmp32dll\x86_64-gf2m.asm(1) : error A2088: END directive required at end of file
Microsoft (R) Macro Assembler (x64) Version 8.00.50727.762
Copyright (C) Microsoft Corporation.  All rights reserved.

NMAKE : fatal error U1077: '"C:\Program Files (x86)\Microsoft Visual Studio 
8\VC\BIN\x86_amd64\ml64.EXE"' : return code '0x1'
Stop.

The build finishes successfuly when nmake is invoked again. The failure is much 
less probable in stand-alone OpenSSL builds.

Used software:

* Windows 7 Professional SP1, 64 bit
* Microsoft Visual Studio 2010 10.0.40219.1 SP1Rel command prompt 
(vcvarsall.bat amd64)
* Perl v5.16.2, Binary build 1602 [296513] provided by ActiveState 
http://www.ActiveState.com

Workaround:

---
--- util/pl/VC-32.plMon Feb 11 16:26:04 2013
+++ util/pl/workaroud_VC-32.pl  Fri Mar  1 16:21:19 2013
@@ -58,6 +58,7 @@
 \$(TMP_D)$o$bname.asm: $source
set ASM=\$(ASM)
\$(PERL) $source \$\@
+   sleep 1
 
 $target: \$(TMP_D)$o$bname.asm
\$(ASM) $afile\$\@ \$(TMP_D)$o$bname.asm
---

I think that the proper solution would be to use temporary files instead of 
pipes when generating ASM on Windows.

Note that the bug is similar to bug #2817 but the workaround described in the 
bug (disable STDOUT buffering) does not work for me.

__
OpenSSL Project http://www.openssl.org
Development Mailing List   openssl-dev@openssl.org
Automated List Manager   majord...@openssl.org