apply in xen/arch/x86/boot
On Wed, Feb 23, 2005 at 01:37:32PM -0500, Jimi Xenidis wrote:> apply in xen/arch/x86/boot> + if ( (in32_ehdr.e_type >> 8) == ET_EXEC && > + (in32_ehdr.e_type & 0xff) != ET_EXEC ) > + { > + swap = 1; > + swap_ehdr32(&in32_ehdr); > + } > + else if ( (in32_ehdr.e_type >> 8) != ET_EXEC && > + (in32_ehdr.e_type & 0xff) == ET_EXEC )This looks wrong. The endianess is figured by e.ident[EI_DATA] not by reading e_type to see if we can recognize the type in all endianess. Since you didn''t remove the test just before if ( !IS_ELF(in32_ehdr) || (in32_ehdr.e_ident[EI_DATA] != ELFDATA2LSB) ) { fprintf(stderr, "Input image must be a little-endian Elf image.\n"); return 1; } so you can''t be there with a valid MSB ELF image. the only explanation I have for the patch, is that you have an ELF file that report to be LSB but is in fact MSB ... what''s the point to allow broken ELF file ? please explain why this patch is necessary, thanks, -- Vincent Hanquez ------------------------------------------------------- SF email is sponsored by - The IT Product Guide Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now. http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click _______________________________________________ Xen-devel mailing list Xen-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/xen-devel
>>>>> "VH" == Vincent Hanquez <tab@snarc.org> writes:VH> On Wed, Feb 23, 2005 at 01:37:32PM -0500, Jimi Xenidis wrote: >> apply in xen/arch/x86/boot >> + if ( (in32_ehdr.e_type >> 8) == ET_EXEC && >> + (in32_ehdr.e_type & 0xff) != ET_EXEC ) >> + { >> + swap = 1; >> + swap_ehdr32(&in32_ehdr); >> + } >> + else if ( (in32_ehdr.e_type >> 8) != ET_EXEC && >> + (in32_ehdr.e_type & 0xff) == ET_EXEC ) VH> This looks wrong. The endianess is figured by e.ident[EI_DATA] not by VH> reading e_type to see if we can recognize the type in all endianess. This test is discover if the build machine is the same endian of the image or not. VH> Since you didn''t remove the test just before VH> if ( !IS_ELF(in32_ehdr) || VH> (in32_ehdr.e_ident[EI_DATA] != ELFDATA2LSB) ) VH> { VH> fprintf(stderr, "Input image must be a little-endian Elf image.\n"); VH> return 1; VH> } VH> so you can''t be there with a valid MSB ELF image. This is a different test, AFAIK this program is not relavent for any MSB ELF image. If we find an MSB ELF that does require this "massaging" then we can revisit this last test. VH> the only explanation I have for the patch, is that you have an ELF file VH> that report to be LSB but is in fact MSB ... This patch is to solve the problem of running the program on then MSB build machine that is operating on the LSB ELF image. VH> please explain why this patch is necessary, thanks, I think I''ve addressed your issues. -JX ------------------------------------------------------- SF email is sponsored by - The IT Product Guide Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now. http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click _______________________________________________ Xen-devel mailing list Xen-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/xen-devel
On Thu, Feb 24, 2005 at 07:31:04AM -0500, Jimi Xenidis wrote:> VH> This looks wrong. The endianess is figured by e.ident[EI_DATA] not by > VH> reading e_type to see if we can recognize the type in all endianess. > > This test is discover if the build machine is the same endian of the > image or not.make sense. I assumed, because of x86 directory in the path, that only cross-compiler would compile that, but indeed that''s just a tool. sorry for the noise :) -- Vincent Hanquez ------------------------------------------------------- SF email is sponsored by - The IT Product Guide Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now. http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click _______________________________________________ Xen-devel mailing list Xen-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/xen-devel