davidsen@steinmetz.ge.com (William E. Davidsen Jr) (03/14/89)
Are there any *standard* caveats about bitfield initialization? I have
read dpANS 3.5.2.1 (bitfields) and 3.5.7 (initialization) without
finding any, but I had one of my students bring me a simple program
which fails on two compilers, and I want to be sure that I'm not missing
something before I tell him that both compilers are broken.
I have already added this to my personal list of things to avoid in a
program I have to port to a number of machines...
Mail or post depending on your perception of public interest, if you can
point out a section I've missed, feel free.
--
bill davidsen (wedu@ge-crd.arpa)
{uunet | philabs}!steinmetz!crdos1!davidsen
"Stupidity, like virtue, is its own reward" -mehenry@utzoo.uucp (Henry Spencer) (03/16/89)
In article <13363@steinmetz.ge.com> davidsen@crdos1.UUCP (bill davidsen) writes: >Are there any *standard* caveats about bitfield initialization? ... One: many old compilers won't do it, or don't do it correctly. -- Welcome to Mars! Your | Henry Spencer at U of Toronto Zoology passport and visa, comrade? | uunet!attcan!utzoo!henry henry@zoo.toronto.edu
afscian@violet.waterloo.edu (Anthony Scian) (03/17/89)
In article <1989Mar15.165019.21562@utzoo.uucp> henry@utzoo.uucp (Henry Spencer) writes: >In article <13363@steinmetz.ge.com> davidsen@crdos1.UUCP (bill davidsen) writes: >>Are there any *standard* caveats about bitfield initialization? ... >One: many old compilers won't do it, or don't do it correctly. What does the standard say about this? Is there a correct ANSI standard method? I don't recall seeing anything in the standard detailing this. //// Anthony Scian afscian@violet.uwaterloo.ca afscian@violet.waterloo.edu //// "I can't believe the news today, I can't close my eyes and make it go away" -U2
karl@haddock.ima.isc.com (Karl Heuer) (03/17/89)
In <1989Mar15.165019.21562@utzoo.uucp> utzoo!henry (Henry Spencer) writes: >In <13363@steinmetz.ge.com> crdos1!davidsen (bill davidsen) writes: >>Are there any *standard* caveats about bitfield initialization? ... > >One: many old compilers won't do it, or don't do it correctly. Since Bill posted to comp.std.c and referenced the (d)pANS, I presume that old (in the sense of pre-ANSI) compilers are not the issue. On the other hand, since he didn't post some sample code and output, I'm also not sure what the problem is. The only such bitfield confusion I know of offhand concerns the rules for structs that have unnamed bitfields. Is that what we're talking about? Karl W. Z. Heuer (ima!haddock!karl or karl@haddock.isc.com), The Walking Lint
henry@utzoo.uucp (Henry Spencer) (03/18/89)
In article <12417@watdragon.waterloo.edu> afscian@violet.waterloo.edu (Anthony Scian) writes: >What does the standard say about this? Is there a correct ANSI standard >method? I don't recall seeing anything in the standard detailing this. Precisely. Initialization of bitfields is not special in any way in ANSI C -- they are ordinary structure members for this purpose. -- Welcome to Mars! Your | Henry Spencer at U of Toronto Zoology passport and visa, comrade? | uunet!attcan!utzoo!henry henry@zoo.toronto.edu
karl@haddock.ima.isc.com (Karl Heuer) (03/22/89)
In article <1989Mar17.175208.1064@utzoo.uucp> henry@utzoo.uucp (Henry Spencer) writes: >Initialization of bitfields is not special in any way in ANSI C -- they are >ordinary structure members for this purpose. Except that "All unnamed structure or union members are ignored during initialization" (3.5.7), and I believe that only bitfields can be unnamed. So, what is the correct way to initialize an object of type struct alpha { int a; struct beta { int :3; } b; int c; } , where one of the members is a struct with no initializable members? Does the initializer {1,2} initialize a and c? Or is {1,{},2} the proper (fully- bracketed) syntax? Karl W. Z. Heuer (ima!haddock!karl or karl@haddock.isc.com), The Walking Lint
scjones@sdrc.UUCP (Larry Jones) (03/23/89)
In article <12112@haddock.ima.isc.com>, karl@haddock.ima.isc.com (Karl Heuer) writes: > So, what is the correct way to initialize an object of type > struct alpha { int a; struct beta { int :3; } b; int c; } > , where one of the members is a struct with no initializable members? Does > the initializer {1,2} initialize a and c? Or is {1,{},2} the proper (fully- > bracketed) syntax? The most recent draft adds a caveat to section 3.5.2.1 (Structure and union specifiers) that "If the struct-declaration-list contains no named members, the behavior is undefined." Thus your example structure is non-portable which makes the initialization question moot. ---- Larry Jones UUCP: uunet!sdrc!scjones SDRC scjones@sdrc.UU.NET 2000 Eastman Dr. BIX: ltl Milford, OH 45150 AT&T: (513) 576-2070 "When all else fails, read the directions."