Re: [gentoo-dev] [PATCH v2] sysroot.eclass: Add support for running Windows binaries via Wine
Eli Schwartz writes:
> On 1/27/26 10:01 AM, James Le Cuirot wrote:
>
+ # Assume that Wine can do its own CPU emulation.
+ install -m0755 /dev/stdin "${SCRIPT}" <<-EOF || die
>>>
>>> (What about my '-' question?)
>>>
+ #!/bin/sh
+ SANDBOX_ON=0 LD_PRELOAD=
WINEPATH="\${WINEPATH}\${WINEPATH+;};${winepath//\//\\}" exec wine "\${@}"
>>>
>>> shellcheck sez:
>>> SC3060 (warning): In POSIX sh, string replacement is undefined.
>>
>> I think shellcheck iz wrong. The replacement happens in the Bash context, not
>> the /bin/sh context.
>
>
> I recommend assuming by default that shellcheck is always wrong.
That's what I do. I looked at first, was suspicious, and then ran it.
> But also, install a package and then run shellcheck on the installed file --
> don't even *try* to run shellcheck on strings of shell embedded in other
> shell scripts
.. but I missed that it's evaluated in the bash context instead.
signature.asc
Description: PGP signature
Re: [gentoo-dev] [PATCH v2] sysroot.eclass: Add support for running Windows binaries via Wine
On 1/27/26 10:01 AM, James Le Cuirot wrote:
>>> + # Assume that Wine can do its own CPU emulation.
>>> + install -m0755 /dev/stdin "${SCRIPT}" <<-EOF || die
>>
>> (What about my '-' question?)
>>
>>> + #!/bin/sh
>>> + SANDBOX_ON=0 LD_PRELOAD=
>>> WINEPATH="\${WINEPATH}\${WINEPATH+;};${winepath//\//\\}" exec wine "\${@}"
>>
>> shellcheck sez:
>> SC3060 (warning): In POSIX sh, string replacement is undefined.
>
> I think shellcheck iz wrong. The replacement happens in the Bash context, not
> the /bin/sh context.
I recommend assuming by default that shellcheck is always wrong. But
also, install a package and then run shellcheck on the installed file --
don't even *try* to run shellcheck on strings of shell embedded in other
shell scripts
--
Eli Schwartz
OpenPGP_signature.asc
Description: OpenPGP digital signature
Re: [gentoo-dev] [PATCH v2] sysroot.eclass: Add support for running Windows binaries via Wine
On Tue, Jan 27, 2026 at 02:29:25AM +, Sam James wrote:
> James Le Cuirot writes:
>
> > We do not support installing Wine for other architectures, so we assume
> > that Wine can do its own CPU emulation.
> >
> > The sandbox cannot handle Wine and emits a warning even if you allow all
> > writes, so it needs to be disabled.
> >
> > Unlike Linux shared libraries, Windows DLLs are placed in bin rather
> > than lib, so we need to include that in WINEPATH. lib has been included
> > too, just in case.
> >
> > binfmt_misc supports Windows binaries, but there is no point in
> > supporting that combination here, as Wine would need to be installed
> > locally regardless. This cannot be extended into a container like it can
> > with QEMU.
> >
> > This has been tested with i686-w64-mingw32 from crossdev against
> > dev-vcs/git[iconv] after dropping the iconv cross patch.
> >
> > Signed-off-by: James Le Cuirot
> > ---
> > eclass/sysroot.eclass | 16 +++-
> > 1 file changed, 15 insertions(+), 1 deletion(-)
> >
> > I was quite surprised to find that winepath itself runs under Wine. I don't
> > like that much and thought it was reasonable to convert these particular
> > paths
> > by hand.
> >
> > I also realised that libgcc is important here after finding that zstd.exe
> > would
> > not run without it in the WINEPATH.
> >
> > diff --git a/eclass/sysroot.eclass b/eclass/sysroot.eclass
> > index 0fae642bf593c..3d710f4917419 100644
> > --- a/eclass/sysroot.eclass
> > +++ b/eclass/sysroot.eclass
> > @@ -80,7 +80,21 @@ sysroot_make_run_prefixed() {
> > fi
> > fi
> >
> > - if [[ ${QEMU_ARCH} == $(qemu_arch "${CBUILD}") ]]; then
> > + if [[ ${CHOST} = *-mingw32 ]]; then
> > + if ! type -P wine >/dev/null; then
> > + einfo "Wine not found. Continuing without ${SCRIPT##*/}
> > wrapper."
> > + return 1
> > + fi
> > +
> > + # UNIX paths can work, but programs will not expect this in
> > %PATH%.
> > + local
> > winepath="Z:${LIBGCC};Z:${MYEROOT}/bin;Z:${MYEROOT}/usr/bin;Z:${MYEROOT}/$(get_libdir);Z:${MYEROOT}/usr/$(get_libdir)"
> > +
> > + # Assume that Wine can do its own CPU emulation.
> > + install -m0755 /dev/stdin "${SCRIPT}" <<-EOF || die
>
> (What about my '-' question?)
>
> > + #!/bin/sh
> > + SANDBOX_ON=0 LD_PRELOAD=
> > WINEPATH="\${WINEPATH}\${WINEPATH+;};${winepath//\//\\}" exec wine "\${@}"
>
> shellcheck sez:
> SC3060 (warning): In POSIX sh, string replacement is undefined.
I think shellcheck iz wrong. The replacement happens in the Bash context, not
the /bin/sh context.
>
> > + EOF
> > + elif [[ ${QEMU_ARCH} == $(qemu_arch "${CBUILD}") ]]; then
> > # glibc: ld.so is a symlink, ldd is a binary.
> > # musl: ld.so doesn't exist, ldd is a symlink.
> > local DLINKER candidate
Re: [gentoo-dev] [PATCH v2] sysroot.eclass: Add support for running Windows binaries via Wine
Sam James writes:
> James Le Cuirot writes:
>
>> We do not support installing Wine for other architectures, so we assume
>> that Wine can do its own CPU emulation.
>>
>> The sandbox cannot handle Wine and emits a warning even if you allow all
>> writes, so it needs to be disabled.
>>
>> Unlike Linux shared libraries, Windows DLLs are placed in bin rather
>> than lib, so we need to include that in WINEPATH. lib has been included
>> too, just in case.
>>
>> binfmt_misc supports Windows binaries, but there is no point in
>> supporting that combination here, as Wine would need to be installed
>> locally regardless. This cannot be extended into a container like it can
>> with QEMU.
>>
>> This has been tested with i686-w64-mingw32 from crossdev against
>> dev-vcs/git[iconv] after dropping the iconv cross patch.
>>
>> Signed-off-by: James Le Cuirot
>> ---
>> eclass/sysroot.eclass | 16 +++-
>> 1 file changed, 15 insertions(+), 1 deletion(-)
>>
>> I was quite surprised to find that winepath itself runs under Wine. I don't
>> like that much and thought it was reasonable to convert these particular
>> paths
>> by hand.
>>
>> I also realised that libgcc is important here after finding that zstd.exe
>> would
>> not run without it in the WINEPATH.
>>
>> diff --git a/eclass/sysroot.eclass b/eclass/sysroot.eclass
>> index 0fae642bf593c..3d710f4917419 100644
>> --- a/eclass/sysroot.eclass
>> +++ b/eclass/sysroot.eclass
>> @@ -80,7 +80,21 @@ sysroot_make_run_prefixed() {
>> fi
>> fi
>>
>> -if [[ ${QEMU_ARCH} == $(qemu_arch "${CBUILD}") ]]; then
>> +if [[ ${CHOST} = *-mingw32 ]]; then
>> +if ! type -P wine >/dev/null; then
>> +einfo "Wine not found. Continuing without ${SCRIPT##*/}
>> wrapper."
>> +return 1
>> +fi
>> +
>> +# UNIX paths can work, but programs will not expect this in
>> %PATH%.
>> +local
>> winepath="Z:${LIBGCC};Z:${MYEROOT}/bin;Z:${MYEROOT}/usr/bin;Z:${MYEROOT}/$(get_libdir);Z:${MYEROOT}/usr/$(get_libdir)"
>> +
>> +# Assume that Wine can do its own CPU emulation.
>> +install -m0755 /dev/stdin "${SCRIPT}" <<-EOF || die
>
> (What about my '-' question?)
(Sorry, missed the answer!)
>
>> +#!/bin/sh
>> +SANDBOX_ON=0 LD_PRELOAD=
>> WINEPATH="\${WINEPATH}\${WINEPATH+;};${winepath//\//\\}" exec wine "\${@}"
>
> shellcheck sez:
> SC3060 (warning): In POSIX sh, string replacement is undefined.
>
>> +EOF
>> +elif [[ ${QEMU_ARCH} == $(qemu_arch "${CBUILD}") ]]; then
>> # glibc: ld.so is a symlink, ldd is a binary.
>> # musl: ld.so doesn't exist, ldd is a symlink.
>> local DLINKER candidate
signature.asc
Description: PGP signature
Re: [gentoo-dev] [PATCH v2] sysroot.eclass: Add support for running Windows binaries via Wine
James Le Cuirot writes:
> We do not support installing Wine for other architectures, so we assume
> that Wine can do its own CPU emulation.
>
> The sandbox cannot handle Wine and emits a warning even if you allow all
> writes, so it needs to be disabled.
>
> Unlike Linux shared libraries, Windows DLLs are placed in bin rather
> than lib, so we need to include that in WINEPATH. lib has been included
> too, just in case.
>
> binfmt_misc supports Windows binaries, but there is no point in
> supporting that combination here, as Wine would need to be installed
> locally regardless. This cannot be extended into a container like it can
> with QEMU.
>
> This has been tested with i686-w64-mingw32 from crossdev against
> dev-vcs/git[iconv] after dropping the iconv cross patch.
>
> Signed-off-by: James Le Cuirot
> ---
> eclass/sysroot.eclass | 16 +++-
> 1 file changed, 15 insertions(+), 1 deletion(-)
>
> I was quite surprised to find that winepath itself runs under Wine. I don't
> like that much and thought it was reasonable to convert these particular paths
> by hand.
>
> I also realised that libgcc is important here after finding that zstd.exe
> would
> not run without it in the WINEPATH.
>
> diff --git a/eclass/sysroot.eclass b/eclass/sysroot.eclass
> index 0fae642bf593c..3d710f4917419 100644
> --- a/eclass/sysroot.eclass
> +++ b/eclass/sysroot.eclass
> @@ -80,7 +80,21 @@ sysroot_make_run_prefixed() {
> fi
> fi
>
> - if [[ ${QEMU_ARCH} == $(qemu_arch "${CBUILD}") ]]; then
> + if [[ ${CHOST} = *-mingw32 ]]; then
> + if ! type -P wine >/dev/null; then
> + einfo "Wine not found. Continuing without ${SCRIPT##*/}
> wrapper."
> + return 1
> + fi
> +
> + # UNIX paths can work, but programs will not expect this in
> %PATH%.
> + local
> winepath="Z:${LIBGCC};Z:${MYEROOT}/bin;Z:${MYEROOT}/usr/bin;Z:${MYEROOT}/$(get_libdir);Z:${MYEROOT}/usr/$(get_libdir)"
> +
> + # Assume that Wine can do its own CPU emulation.
> + install -m0755 /dev/stdin "${SCRIPT}" <<-EOF || die
(What about my '-' question?)
> + #!/bin/sh
> + SANDBOX_ON=0 LD_PRELOAD=
> WINEPATH="\${WINEPATH}\${WINEPATH+;};${winepath//\//\\}" exec wine "\${@}"
shellcheck sez:
SC3060 (warning): In POSIX sh, string replacement is undefined.
> + EOF
> + elif [[ ${QEMU_ARCH} == $(qemu_arch "${CBUILD}") ]]; then
> # glibc: ld.so is a symlink, ldd is a binary.
> # musl: ld.so doesn't exist, ldd is a symlink.
> local DLINKER candidate
signature.asc
Description: PGP signature
