How to deal with unrecognizable RTL code

Hello, guys: The porting is really a difficult and huge job. So many things I don't know or miss result in countless bugs. I'd like to thank you for your guys in maillist helping so much, especially, Ian Lance Taylor. Here I encounter some unrecognizable RTL (R0 to R1

Re: gcj build issues.

Dave Korn wrote: > Edward Peschko wrote: > >> 3. ecj not part of the build, hence causing at runtime: >> >> ld.so.1: ecj1: fatal: libgcc_s.so.1: version `GCC_4.2.0' not >> found (required by file >> /userdata/ebay/interface/FI/tools/beta/lib/libgcj.so.10) >> ld.so.1: ecj1: fa

Re: How to deal with unrecognizable RTL code

田晓南 wrote: > Hello, guys: > The porting is really a difficult and huge job. So many things I > don't know or miss result in countless bugs. It should not be hard. You have to tell us however why this is unrecognizable, that is, what would be the "closest" recognizable insn supported by you

Re: i370 port

Paul Edwards wrote: > But sometimes r_or_s_operand is being used as a source, in > which case, the constant is fine. But when it is used as a > destination, it is not fine. > > What is the *simplest* way of changing the setup so that the > code generation remains the same, but the warning is eli

Re: gcj build issues.

> "Edward" == Edward Peschko writes: Edward> 2. hardcoded '/bin/sh' references in files cause build Edward> incompatibilities (fails on solaris - Edward> Comparing stage 2 to stage 3.. since they are hardcoded, Edward> not fixable by setting Edward> CONFIG_SHELL)

[plugins] language makefiles and installing cp/ headers

Hi, I tried to make my plugin compile using only installed headers and turned out that I needed more gcc headers to be installed. Unfortunately, I needed C++ headers which currently aren't installed. I modified the cp/Make-lang.in with logic copied from gcc/Makefile.in to add a c++.install-plu

Re: [plugins] language makefiles and installing cp/ headers

On Fri, 19 Jun 2009, Taras Glek wrote: > Hi, > I tried to make my plugin compile using only installed headers and turned out > that I needed more gcc headers to be installed. Unfortunately, I needed C++ > headers which currently aren't installed. I modified the cp/Make-lang.in with > logic copied

Re: git mirror at gcc.gnu.org

On 06/16/2009 03:51 PM, Bernie Innocenti wrote: Yes, but which one should die? Will cloners get confused by a repository where the master branch is missing? clone will get whatever branch is currently active in the cloned repository, doesn't matter what it's called. Currently master is out

Should -Wjump-misses-init be in -Wall?

I recently added the new -Wjump-misses-init warning option. It warns when a goto or switch jumps into the scope of an initialized variable without actually initializing the variable. I added the warning to -Wall because it seems to me to fit the criteria of -Wall: a dubious code practice which is

Re: Should -Wjump-misses-init be in -Wall?

Ian Lance Taylor wrote: I recently added the new -Wjump-misses-init warning option. It warns when a goto or switch jumps into the scope of an initialized variable without actually initializing the variable. I added the warning to -Wall because it seems to me to fit the criteria of -Wall: a dubi

Re: Should -Wjump-misses-init be in -Wall?

Hello Ian, out of curiosity: * Ian Lance Taylor wrote on Sat, Jun 20, 2009 at 07:12:50AM CEST: > Any opinions on this? Should I take the new warning out of -Wall? Is the missing of an initialization detected elsewhere, or can it be detected elsewhere, maybe only in cases where it actually leads