|
From: | Paolo Bonzini |
Subject: | Re: Fwd: sed porting trouble |
Date: | Fri, 01 Oct 2010 15:25:03 +0200 |
User-agent: | Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.2.9) Gecko/20100907 Fedora/3.1.3-1.fc13 Lightning/1.0b3pre Mnenhy/0.8.3 Thunderbird/3.1.3 |
On 10/01/2010 03:10 PM, Schmitz, Joachim wrote:
Hi Paolo In general the entire documentation for thie platform is here: http://www.hp.com/go/nonstop-docs, In this case you'd probably want to look at the (currently) latest version of the C/C++ Programmer's Guide http://bizsupport1.austin.hp.com/bc/docs/support/SupportManual/c02128447/c02128447.pdf , Chapter 13 Compiler Pragmas. Hmm, reading is of advantage sometime, I just see that instead of the pragma, I can use a compiler switch -Wcolumns=c, so I should be able to get around that issue without those 2 patches, by just tweaking my cc wrapper script slightly (i.e. adding a -Wcolumns=32767, the maximum, to the switches it uses to call the real c89/c99 C/C++ compiler). (Testing....) Rats! It doesn't work! Seems the Compiler switch does not precedence over the pragma.
Here it is (page 188):"At each #include directive, each included file initially assumes the last-column value in effect when the #include directive appeared. If a COLUMNS pragma is specified in the included file, the compiler uses the last-column value specified."
This basically means that the NonStop system includes (by enforcing #pragma column 80) only expect to include other NonStop system includes. This is clearly not the case in gnulib. So, it looks like we must make sure that the system-header replacements have 80-columns lines. Sigh.
Paolo
[Prev in Thread] | Current Thread | [Next in Thread] |