Re: [PATCH] diffcore.h: be explicit about the signedness of is_binary

2014-02-24 Thread Junio C Hamano
Richard Lowe writes: > Bitfields need to specify their signedness explicitly or the compiler is > free to default as it sees fit. With compilers that default 'unsigned' > (SUNWspro 12 seems to do this) the tri-state nature of is_binary > vanishes and all files are treated as binary. > > Signed-o

Re: [PATCH] diffcore.h: be explicit about the signedness of is_binary

2014-02-23 Thread Jeff King
On Sun, Feb 23, 2014 at 07:54:47PM -0500, Richard Lowe wrote: > Bitfields need to specify their signedness explicitly or the compiler is > free to default as it sees fit. With compilers that default 'unsigned' > (SUNWspro 12 seems to do this) the tri-state nature of is_binary > vanishes and all f

[PATCH] diffcore.h: be explicit about the signedness of is_binary

2014-02-23 Thread Richard Lowe
Bitfields need to specify their signedness explicitly or the compiler is free to default as it sees fit. With compilers that default 'unsigned' (SUNWspro 12 seems to do this) the tri-state nature of is_binary vanishes and all files are treated as binary. Signed-off-by: Richard Lowe --- diffcore