aboutsummaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* dmi: check both the AC and ID flags at the same timeHEADmasterH. Peter Anvin2019-02-202-8/+15
| | | | | | | There is no reason to do the fairly complex flag-test sequence twice, when we might as well check for both flags at the same time. Signed-off-by: H. Peter Anvin <hpa@zytor.com>
* cpu_has_eflag(): mark output registers as earlyclobbersH. Peter Anvin2019-02-201-1/+1
| | | | | | | | | | | We assign values to the output registers before the input operand is used, so we have to use =&r to make sure the input operand isn't clobbered. Note: "ri" is correct; it cannot safely be a memory operand, because the stack pointer isn't where the compiler would expect it to be. Signed-off-by: H. Peter Anvin <hpa@zytor.com>
* cpu_has_eflag(): eliminate if the answer is already knownH. Peter Anvin2019-02-202-1/+20
| | | | | | | There is no need to test for a flag if the answer can be given at compile time. Signed-off-by: H. Peter Anvin <hpa@zytor.com>
* Fix all warnings, and better separate code that should not be mixedsyslinux-6.04-pre3H. Peter Anvin (Intel)2019-02-07262-592/+587
| | | | | | | | | Clean up a number of warnings in the tree. Refactor especially the core code so we don't end up building a bunch of BIOS-specific code for EFI, and general build cleanups. Signed-off-by: H. Peter Anvin (Intel) <hpa@zytor.com>
* Makefile: allow the user to set additional warning flagsH. Peter Anvin (Intel)2019-02-072-2/+3
| | | | | | | Make the variable WARNOPT available for the user to set additional warning options, e.g. to quiet new compiler warnings. Signed-off-by: H. Peter Anvin (Intel) <hpa@zytor.com>
* com32.mk: removeH. Peter Anvin (Intel)2019-02-073-126/+5
| | | | | | | The file com32.mk is no longer used, as we don't use the traditional com32 format anymore (and haven't since version 4.xx). Signed-off-by: H. Peter Anvin (Intel) <hpa@zytor.com>
* gnu-efi: update to version 3.0.9-3-gd34132eH. Peter Anvin (Intel)2019-02-072-11/+15
| | | | | | | | Update gnu-efi to version 3.0.9-3-gd34132e; this is 3.0.9 plus three bugfix patches from the gnu-efi git tree. Suggested-by: Joakim Tjernlund <Joakim.Tjernlund@infinera.com> Signed-off-by: H. Peter Anvin (Intel) <hpa@zytor.com>
* install: don't install obsolete file com32.ldH. Peter Anvin (Intel)2019-02-061-1/+0
| | | | | | | | com32.ld has been obsolete for a long time, and has been removed now; don't install it either. Reported-by: Joakim Tjernlund <Joakim.Tjernlund@infinera.com> Signed-off-by: H. Peter Anvin (Intel) <hpa@zytor.com>
* Merge remote-tracking branch 'joakim/master'H. Peter Anvin (Intel)2019-02-065-14/+4
|\
| * Use CC for CC_FOR_BUILD when unspecified.Joakim Tjernlund2017-11-092-0/+4
| |
| * Update the longjump calls to fit the new declarationBenoît Allard2017-11-091-3/+2
| |
| * The VPrint definition is now part of the exports of gnu-efiBenoît Allard2017-11-091-9/+0
| | | | | | | | Signed-off-by: Benoît Allard <benoit.allard@greenbone.net>
| * New gnu-efi dir in 3.0.6Joakim Tjernlund2017-11-092-2/+2
| |
| * Update gnu-efi submodule to 3.0.6Joakim Tjernlund2017-11-091-0/+0
| |
| * efi/wrapper: build it with the host toolchain.Benoît Allard2017-11-091-1/+1
| | | | | | | | | | | | | | The wrapper program is executed on the build machine, so it should be built with CC_FOR_BUILD. Signed-off-by: Benoît Allard <benoit.allard@greenbone.net>
| * utils: Use the host toolchain to build.Benoît Allard2017-11-091-6/+6
| | | | | | | | | | | | | | The utilities are meant to run on the host machine, hence must be built using the host toolchain. Signed-off-by: Benoît Allard <benoit.allard@greenbone.net>
| * memdisk: Force ld output format to 32-bitsBenoît Allard2017-11-091-1/+1
| | | | | | | | | | | | | | On toolchains where the default output is x86_64, we need to be consistent with the other .o files Signed-off-by: Benoît Allard <benoit.allard@greenbone.net>
* | Makefile.private: don't bother generating .bz2 filesH. Peter Anvin (Intel)2019-02-051-2/+0
| | | | | | | | | | | | | | .gz files for long-term standardization and speed, .xz for small size. .bz2 no longer serves either of these functions. Signed-off-by: H. Peter Anvin (Intel) <hpa@zytor.com>
* | Makefile.private: remove remaining junk from previous processsyslinux-6.04-pre2H. Peter Anvin2019-02-051-5/+2
| | | | | | | | | | | | Don't create a bogus .tar.gz file due to stay rule... Signed-off-by: H. Peter Anvin <hpa@zytor.com>
* | Makefile.private: correct the names of prerelease tarballsH. Peter Anvin2019-02-051-3/+3
| |
* | Makefile.private: generate .tar.bz2 and .tar.xz filesH. Peter Anvin2019-02-051-3/+8
| | | | | | | | | | | | | | Auto-generate .tar.bz2 and .tar.xz files for hosts other than kernel.org. Signed-off-by: H. Peter Anvin <hpa@zytor.com>
* | libpng: update to 1.6.36H. Peter Anvin2019-02-0434-20893/+36640
| | | | | | | | | | | | Update libpng to version 1.6.36 due to errata in earlier versions. Signed-off-by: H. Peter Anvin <hpa@zytor.com>
* | Reorganize and clean up a bunch of the x86 codeH. Peter Anvin2019-02-0446-612/+422
| | | | | | | | | | | | | | We were doing a bunch of i386-specific things even on x86-64. Fix this, and merge x86 definitions where possible. Signed-off-by: H. Peter Anvin <hpa@zytor.com>
* | com32: remove setjmp.S stub fileH. Peter Anvin2019-02-042-9/+2
| | | | | | | | | | | | Compile setjmp.S directly out of the $(ARCH) directory. Signed-off-by: H. Peter Anvin <hpa@zytor.com>
* | libcc: move libgcc files into arch-specific directoriesH. Peter Anvin2019-02-0411-8/+3
| | | | | | | | | | | | | | | | We would compile these for for x86-64 even though they are i386-specific; the resulting code was meaningless but nevertheless included in the output. Signed-off-by: H. Peter Anvin <hpa@zytor.com>
* | com32/lib: clean up unused filesH. Peter Anvin2019-02-046-613/+2
| | | | | | | | | | | | Clean up obsolete unused files; simplify setjmp.S stub Signed-off-by: H. Peter Anvin <hpa@zytor.com>
* | zlib: update to version 1.2.11H. Peter Anvin2019-01-2822-1280/+2308
| | | | | | | | | | | | Update zlib to version 1.2.11 due to errata in previous versions. Signed-off-by: H. Peter Anvin <hpa@zytor.com>
* | core/Makefile: sort to be reproducibleLukas Schwaighofer2018-10-251-5/+5
| | | | | | | | Signed-off-by: Lukas Schwaighofer <lukas@schwaighofer.name>
* | ISOLINUX: Fix checksum calculation in lzo/prepcore.cThomas Schmitt2018-10-251-1/+1
| | | | | | | | | | | | | | | | | | | | | | The prescription for Boot Info Table says that checksumming begins at byte 64 of isolinux.bin. When prepcore writes isolinux.bin it begins copying bytes from the input file at the offset given by variable "start". But it begins checksumming at offset 64 of the input file. The problem exists since introduction of prepcore by release 4.00. ISO 9660 programs usually fix it when they write the Boot Info Table. Signed-off-by: Thomas Schmitt <scdbackup@gmx.net>
* | efi/wrapper: build it with the host toolchain.Benoît Allard2017-11-231-1/+1
| | | | | | | | | | | | | | The wrapper program is executed on the build machine, so it should be built with CC_FOR_BUILD. Signed-off-by: Benoît Allard <benoit.allard@greenbone.net>
* | utils: Use the host toolchain to build.Benoît Allard2017-11-231-6/+6
| | | | | | | | | | | | | | | | | | | | | | The utilities are meant to run on the host machine, hence must be built using the host toolchain. Signed-off-by: Benoît Allard <benoit.allard@greenbone.net> (GC): Minor merge conflict Signed-off-by: Gene Cumm <gene.cumm@gmail.com>
* | Use CC for CC_FOR_BUILD when unspecified.Joakim Tjernlund2017-11-232-0/+4
| | | | | | | | | | | | (GC): Let's put the pre-requisite first rather than making a FTBFS commit Signed-off-by: Gene Cumm <gene.cumm@gmail.com>
* | memdisk: Force ld output format to 32-bitsBenoît Allard2017-11-231-1/+1
|/ | | | | | | On toolchains where the default output is x86_64, we need to be consistent with the other .o files Signed-off-by: Benoît Allard <benoit.allard@greenbone.net>
* utils: use Digest::SHA for sha1passSalvatore Bonaccorso2017-11-041-2/+2
| | | | | | | | | | Debian (Closes: #625812) Author: Salvatore Bonaccorso <carnil@debian.org> Logic can be added if we need to check between these. Signed-off-by: Gene Cumm <gene.cumm@gmail.com>
* libinstaller: Use SOURCE_DATE_EPOCH for synthesized modification time stampsLukas Schwaighofer2017-11-041-3/+6
| | | | | | | | Patches bin2c.pl to use the value of SOURCE_DATE_EPOCH instead of real modification time stamp (only if SOURCE_DATE_EPOCH is set). Author: Lukas Schwaighofer <lukas@schwaighofer.name> Signed-off-by: Gene Cumm <gene.cumm@gmail.com>
* make: Sort object files for reproducible linking orderReiner Herrmann2017-11-041-4/+4
| | | | | | | | | Author: Reiner Herrmann <reiner@reiner-h.de> Change be645d7a was not extended to the efi/Makefile previously; this patch addresses this. Signed-off-by: Gene Cumm <gene.cumm@gmail.com>
* man: Updating list of supported filesystems in extlinuxDaniel Baumann2017-11-041-7/+16
| | | | | | | Debian (Closes: #692844); also update contact info. Author: Daniel Baumann <mail@daniel-baumann.ch> Signed-off-by: Gene Cumm <gene.cumm@gmail.com>
* ext4: Fix 64bit featurePaulo Alcantara2017-10-152-9/+15
| | | | | | | | | | | | | | | | | | | | | | As per ext4 specification: > In ext2, ext3, and ext4 (when the 64bit feature is not enabled), the > block group descriptor was only 32 bytes long and therefore ends at > bg_checksum. On an ext4 filesystem with the 64bit feature enabled, the > block group descriptor expands to at least the 64 bytes described below; > the size is stored in the superblock. Since block group descriptor has been expanded to 64 bytes long (when 64 bit feature is enabled), we cannot index ext2_group_desc and return it *directly* -- as we did it in ext2_get_group_desc -- it's still 32 bytes long. Instead, use s_desc_size field from superblock to correctly index and return block group descriptors. Cc: H. Peter Anvin <hpa@zytor.com> Cc: Gene Cumm <gene.cumm@gmail.com> Signed-off-by: Paulo Alcantara <pcacjr@zytor.com>
* efi/udp: Removing the udp_reader efi_bindingJulien Viard de Galbert2017-05-311-22/+2
| | | | Signed-off-by: Julien Viard de Galbert <jviarddegalbert@online.net>
* efi/udp: Add retry disabling UseDefaultAddressJulien Viard de Galbert2017-05-311-0/+12
| | | | | | | | Add retry disabling UseDefaultAddress in core_udp_connect() and core_udp_sendto() Signed-off-by: Julien Viard de Galbert <jviarddegalbert@online.net> Signed-off-by: Gene Cumm <gene.cumm@gmail.com>
* efi/udp: core_udp_connect: use SubnetMask not StationAddress for netmaskJulien Viard de Galbert2017-05-311-1/+1
| | | | | Signed-off-by: Julien Viard de Galbert <jviarddegalbert@online.net> Signed-off-by: Gene Cumm <gene.cumm@gmail.com>
* isohybrid: Open ISO file in binary modeColin Finck2017-05-141-1/+1
| | | | | | Open ISO file in binary mode to ensure that line endings stay untouched. Signed-off-by: Gene Cumm <gene.cumm@gmail.com>
* mbr/isohdpfx.S: Clear CX on INT 13h AH 41h failureMartin Str|mberg2017-03-281-1/+3
| | | | | | If anything goes wrong, clear CX in case it contains garbage. Signed-off-by: Gene Cumm <gene.cumm@gmail.com>
* mbr/isohdpfx.S: correct stack for heads/sectors; revertMartin Str|mberg2017-03-281-3/+3
| | | | | | | | Heads and sectors were pushed in reverse order per isolinux.asm bb519a95 reversed the order of heads/sectors on the stack Signed-off-by: Gene Cumm <gene.cumm@gmail.com>
* Revert "mbr/isohdpfx.S: correct pointer for heads/sectors"Martin Str|mberg2017-03-281-2/+2
| | | | | | This reverts commit 48e94f4fa7b3c32cbd43b6e57c64bc933f76d059. Signed-off-by: Gene Cumm <gene.cumm@gmail.com>
* mbr/isohdpfx.S: correct pointer for heads/sectorsMartin Str|mberg2017-03-251-2/+2
| | | | | | sectors is pushed first then heads but pointers were in reverse order. Signed-off-by: Gene Cumm <gene.cumm@gmail.com>
* pxe: Fix recognition of keeppxe optionAdam Goldman2017-03-052-9/+11
| | | | | | | | | | | | | | | The preprocessor directive IS_PXELINUX is no longer valid. Submitted-by: Adam Goldman <adamg@pobox.com> Move "keeppxe" processing from com32/elflink/ldlinux/kernel.c:new_linux_kernel() to com32/lib/syslinux/load_linux.c:bios_boot_linux() using strstr() (not find_argument in original patch). Should also be processed by firmware->boot_linux() if not yet done. Signed-off-by: Gene Cumm <gene.cumm@gmail.com>
* chain.c32: update reactos= option for modern valueAdy2017-03-042-10/+4
| | | | | | | | | | | | | | | | | | | | Changed from "seg=0:0x8000:0x8100" to "seg=0x0F80" Reactos is "alpha", so the only version of Reactos that its devs support is the "latest" at each time. Older versions of Reactos are not relevant / supported by Reactos' devs. Submitted-By: Ady <ady-sf@hotmail.com> Old 2011-06-14 https://git.reactos.org/?p=reactos.git;a=blob;f=reactos/boot/freeldr/freeldr/include/arch/pc/x86common.h;hb=28e58e6d01892c1f2f0e1d323745e6463cb9e6c9 https://svn.reactos.org/svn/reactos/trunk/reactos/boot/freeldr/freeldr/include/arch/pc/x86common.h?revision=52226&view=markup New 2011-06-18 https://git.reactos.org/?p=reactos.git;a=blob;f=reactos/boot/freeldr/freeldr/include/arch/pc/x86common.h;hb=0eb6daaf66127de803d36d45a4eea29d30a7ca36 https://svn.reactos.org/svn/reactos/trunk/reactos/boot/freeldr/freeldr/include/arch/pc/x86common.h?revision=52346&view=markup Signed-off-by: Gene Cumm <gene.cumm@gmail.com>
* pxe_api.h: finish revert of 9acbffdGene Cumm2016-11-271-3/+3
| | | | | | | | 9acbffd33b9200ffe37833463b4d4478e824295a caused more issues than it solves. There appears to be a bigger underlying issue that the __weak argument masks. Until the underlying issue is solved, we need __weak. Signed-off-by: Gene Cumm <gene.cumm@gmail.com>
* core/legacynet/dnsresolv.c: weaken pxe_dns()Gene Cumm2016-11-271-1/+1
| | | | Signed-off-by: Gene Cumm <gene.cumm@gmail.com>