Date: Fri, 11 Jul 2008 23:14:50 +0300 From: Kostik Belousov <kostikbel@gmail.com> To: Marcin Cieslak <saper@system.pl> Cc: freebsd-emulation@freebsd.org Subject: Re: linux emulation: Preliminary support for more auxvec's [patch] Message-ID: <20080711201450.GB17123@deviant.kiev.zoral.com.ua> In-Reply-To: <4877BAB8.1030804@system.pl> References: <g57h1u$5od$1@ger.gmane.org> <20080711115436.GZ17123@deviant.kiev.zoral.com.ua> <4877BAB8.1030804@system.pl>
next in thread | previous in thread | raw e-mail | index | archive | help
--6xO04bqTG+NC5D6o Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Fri, Jul 11, 2008 at 09:55:36PM +0200, Marcin Cieslak wrote: > Kostik Belousov wrote: > >On Fri, Jul 11, 2008 at 01:43:55PM +0200, Marcin Cieslak wrote: > >>Hello, > >> > >>Attached please find a simple diff to implement additional loader=20 > >>information (for background please see:=20 > >>http://lists.freebsd.org/pipermail/freebsd-emulation/2006-September/002= 591.html) > >Namely, AT_EXECPATH (somewhat similar to the Solaris auxvec of the same > >name) provides (might be relative) path for the binary being executed. >=20 > Thank you very much. >=20 > I have implemented the AT_PLATFORM, therefore all 2.6.16 (and later)=20 > i386/amd64 auxvecs are implemented, except for AT_SYSINFO and=20 > AT_SYSINFO_EHDR that are i386-specific and provide optional way of=20 > invoking linux syscalls (using so-called virtual dynamic shared object). >=20 > I think we don't need those. Do you know of any situation where we _must_ have those new auxvec you implemented ? >=20 > The code is completely untested on i386. My first attempts show that=20 > skype and acroread8 launch faster (probably due to the "hz" effect). >=20 > The patch is here: >=20 > http://akson.sgh.waw.pl/~saper/FreeBSD/linux/auxvec.diff I only briefly looked over it. I suggest that the new AT_COUNT requires more thinking. Look at the src/libexec/rtld-elf/rtld.c, _rtld(), at line 338, where the auxvec is digested by FreeBSD dynamic linker. It seems to be harmless, but at least this is inconsistent. May be, AT_LINUX_PLATFORM etc defines would make more sense. Do you actually need imgp->machine member ? It seems that it is used only inside linux_copyout_strings ? >=20 > This was made against 7-STABLE, but there no major differences in=20 > -current. It is also trivial to port to 64-bit amd64 emulation. Hmm, what are you referencing there ? I know only about one mostly stalled effort. --6xO04bqTG+NC5D6o Content-Type: application/pgp-signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.9 (FreeBSD) iEYEARECAAYFAkh3vzkACgkQC3+MBN1Mb4gcGwCeN8opVdZf5Kw3Pohrs5uQTEOb alUAnj+4JipcTqkYr7ADOFVWaZXBQyKs =cP19 -----END PGP SIGNATURE----- --6xO04bqTG+NC5D6o--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20080711201450.GB17123>