similar to: Two-stage build w/ ninja - Tests still use stage1 compiler / linker ?

Hi, I recently ran into an issue where the dependencies generated in modules.dep created a circular reference, which meant that qemu spun forever in modules_load_dependencies(). I did think about detecting this kind of circular dependency at modules.dep generation time but if we ever move to using DT_NEEDED to track dependencies we would need this functionality in the elflink core anyway.

Folks, I did read the documentation but I am at a loss still. I do have the following: Pass *createFooPass (char *foo1) { return new FooPass (foo1); } static RegisterPass<FooPass> X("foo-pass", "foo", false, true); Where exactly do I supply the non-default crore in this example ? Thanks /Dirk

I try to compile LLVM, clang, lld and libcxx together in one source directory following the instructions on the website. I'm running Arch Linux and already have LLVM, clang, lld and libcxx installed (prebuilt against libstdc++). Now I want a "pure" llvm toolchain. The following options were used: -DBOOTSTRAP_LLVM_ENABLE_LLD -DCMAKE_C_COMPILER=clang -DCMAKE_CXX_COMPILER=clang++

compiler-rt libs must be built with /MT, so the MSVS build is doing the wrong thing here. 2014-10-23 12:52 GMT-07:00 Aaron Ballman <aaron at aaronballman.com>: > On Thu, Oct 23, 2014 at 3:42 PM, Aaron Ballman <aaron at aaronballman.com> wrote: >> On Thu, Oct 23, 2014 at 3:38 PM, Aaron Ballman <aaron at aaronballman.com> wrote: >>> On Thu, Oct 23, 2014 at 2:57

On Sep 30, 2013, at 10:11 PM, Dirk Schroetter <dschroetter at googlemail.com> wrote: > Folks, > > I did read the documentation but I am at a loss still. > > I do have the following: > > Pass *createFooPass (char *foo1) { > return new FooPass (foo1); > } > > static RegisterPass<FooPass> X("foo-pass", "foo", false, true); >

I can build llvm and associated tools in a 1 shot go using gcc and g++ but trying to use only gcc and bootstrap clang to build and link with libc++ and libc++abi on the second phase of the build, the compilation fails I have svn cloned LLVM CLANG CLANG_XTRA_TOOLS LLD LLDB POLLY COMPILER_RT OPENMP LIBCXX LIBCXXABI cmake command below cmake -DCMAKE_BUILD_TYPE=Release -DLLVM_ENABLE_ASSERTIONS=OFF

You also need -DLIBUNWIND_USE_COMPILER_RT=ON -DLIBCXXABI_USE_LLVM_UNWINDER=ON. This is because you set -DLIBCXXABI_USE_COMPILER_RT=ON so libc++abi.so is no longer being linked agains libgcc_s.so which is where the unwinder otherwise comes from. On Sun, May 6, 2018 at 4:05 AM via llvm-dev <llvm-dev at lists.llvm.org> wrote: > I try to compile LLVM, clang, lld and libcxx together in one

Hello, I have a working GCC toolchain and wanted to use LLVM tooling instead. Following the instructions on the website I set up my build directory and have lld as an internal project under tools/. I'd like to make use of lld in stage 2 of the build and thought -DLLVM_ENABLE_LLD was the correct option to pass. However, adding this options makes the configure step fail, complaining that the

On Thu, Oct 23, 2014 at 3:38 PM, Aaron Ballman <aaron at aaronballman.com> wrote: > On Thu, Oct 23, 2014 at 2:57 PM, Aaron Ballman <aaron at aaronballman.com> wrote: >> On Thu, Oct 23, 2014 at 2:46 PM, Timur Iskhodzhanov <timurrrr at google.com> wrote: >>> 2014-10-23 11:34 GMT-07:00 Aaron Ballman <aaron at aaronballman.com>: >>>> On Thu, Oct

Hi, These patches contain support for some features that are already in Syslinux 4 but weren't working properly on the elflink branch. It's another step closer to feature parity with Syslinux 4. Having to jump through the comboot API for localboot support is less than ideal and I'll eventually fix that, probably when we move a big chunk of code from asm to C. Also, there's a

On Tue, Feb 11, 2020 at 02:23:22PM -0800, Andy Lutomirski wrote: > On Tue, Feb 11, 2020 at 5:53 AM Joerg Roedel <joro at 8bytes.org> wrote: > > +void __init no_ghcb_vc_handler(struct pt_regs *regs) > > Isn't there a second parameter: unsigned long error_code? No, the function gets the error-code from regs->orig_ax. This particular function only needs to check for

On Thu, Mar 19, 2020 at 10:13:15AM +0100, Joerg Roedel wrote: > diff --git a/arch/x86/boot/compressed/idt_handlers_64.S b/arch/x86/boot/compressed/idt_handlers_64.S > index bfb3fc5aa144..67ddafab2943 100644 > --- a/arch/x86/boot/compressed/idt_handlers_64.S > +++ b/arch/x86/boot/compressed/idt_handlers_64.S > @@ -75,3 +75,7 @@ SYM_FUNC_END(\name) > .code64 > >

From: Joerg Roedel <jroedel at suse.de> Add the first handler for #VC exceptions. At stage 1 there is no GHCB yet becaue we might still be on the EFI page table and thus can't map memory unencrypted. The stage 1 handler is limited to the MSR based protocol to talk to the hypervisor and can only support CPUID exit-codes, but that is enough to get to stage 2. Signed-off-by: Joerg Roedel

From: Joerg Roedel <jroedel at suse.de> Add the first handler for #VC exceptions. At stage 1 there is no GHCB yet becaue we might still be on the EFI page table and thus can't map memory unencrypted. The stage 1 handler is limited to the MSR based protocol to talk to the hypervisor and can only support CPUID exit-codes, but that is enough to get to stage 2. Signed-off-by: Joerg Roedel

From: Joerg Roedel <jroedel at suse.de> Add the first handler for #VC exceptions. At stage 1 there is no GHCB yet becaue the kernel might still be running on the EFI page table. The stage 1 handler is limited to the MSR based protocol to talk to the hypervisor and can only support CPUID exit-codes, but that is enough to get to stage 2. Signed-off-by: Joerg Roedel <jroedel at

On Tue, Feb 11, 2020 at 5:53 AM Joerg Roedel <joro at 8bytes.org> wrote: > > From: Joerg Roedel <jroedel at suse.de> > > Add the first handler for #VC exceptions. At stage 1 there is no GHCB > yet becaue we might still be on the EFI page table and thus can't map > memory unencrypted. > > The stage 1 handler is limited to the MSR based protocol to talk to

On Tue, Feb 11, 2020 at 5:53 AM Joerg Roedel <joro at 8bytes.org> wrote: > > From: Joerg Roedel <jroedel at suse.de> > > Add the first handler for #VC exceptions. At stage 1 there is no GHCB > yet becaue we might still be on the EFI page table and thus can't map > memory unencrypted. > > The stage 1 handler is limited to the MSR based protocol to talk to

Building LLVM with Visual Studio is a bit of a pain because the available methods are slow and doesn't make good use of multiprocessor systems. >From now on it is possible to build LLVM+Clang with the usual cmake method but using Ninja, an ultra-fast tool that knows how to take advantage of the availabe execution threads. Ultra-fast is no exageration: with a warm cache, a no-op build of

On Jun 17, 2012, at 7:23 PM, Óscar Fuentes wrote: > OTOH it's my personal > opinion that you may be explaining too much: it is not necessary to > explain, nor even mention, the different available build tools on > Windows. As LLVM and Clang are developer tools, one can suppose that > the > user knows how to run a build. IMO, Mikael's efforts are on-target and very

On Sun, Jun 17, 2012 at 7:30 PM, Mikael Lyngvig <mikael at lyngvig.org> wrote: > Hi Christoph, > > It is a good question even though it is obvious :-)  No, for the last few > days, I have not been able to build Clang with make -j2 or -j4 - the build > always fails on a ranlib operation. > > Also, on Windows, GNU Make seems to be somewhat broken with respect to >