~mirabilos/klibc/master : changes

~mirabilos/klibc/master » Changes from revision 2243

From Revision 2243 to 2224

Rev	Summary	Authors	Tags	Date
2243	[klibc] 2.0.14 released, next version is 2.0.15 [klibc] 2.0.14 released, next version is 2.0.15	Ben Hutchings		8 days ago
2242	[klibc] mips64: Set executable base address to fix link fail... [klibc] mips64: Set executable base address to fix link failure on R6 It appears that at some point the GNU tools started using a different default base address for MIPS64 R6 executables; at least this is what I see with the Debian-provided cross-toolchain. This resulted in klibc.so being too far away from the executable for regular jumps to work. Add an explicit linker option to set the base address to the value we expect. Signed-off-by: Ben Hutchings <ben@decadent.org.uk>	Ben Hutchings		8 days ago
2241	[klibc] ipconfig: align reads of ext for RISC architectures [klibc] ipconfig: align reads of ext for RISC architectures Some RISC architectures such as Sparc64 will busfault when reading data from exts if they are not type aligned. Use memcpy to read the data from exts to ensure type alignment. A similar issue was fixed in dhcpcd here: https://github.com/NetworkConfiguration/dhcpcd/issues/430 https://github.com/ColinMcInnes/dhcpcd/commit/07a5fdd7d34b072e52c316877d14c6d2581cb379 Signed-off-by: Jesse Taube <mr.bossman075@gmail.com> [bwh: Change the comment to not be sparc-specific] Signed-off-by: Ben Hutchings <ben@decadent.org.uk>	Jesse Taube		10 days ago
2240	[klibc] syscalls: Support for inotify_init1() [klibc] syscalls: Support for inotify_init1() Linux 2.6.27 added the inotify_init1() system call, a variant of inotify_init() with a flags parameter added. Generate a wrapper for this system call. Newer architectures support only inotify_init1(), so also add a inotify_init() implementation for them. Signed-off-by: Ben Hutchings <ben@decadent.org.uk>	Ben Hutchings		5 months ago
2239	[klibc] syscalls: Remove support for socketcall() [klibc] syscalls: Remove support for socketcall() All architectures that implement socketcall() now also have separate system calls: - arm: added in 2.6.12 - i386: added in 4.3 - m68k: added in 4.3 - mips, mips64: added in 2.1.44 - parisc: always had both - ppc, ppc64: added in 2.6.37 - s390, s390x: added in 4.3 - sh: added in 2.6.37 - sparc, sparc64: added in 4.4 and 4.20 We can therefore unconditionally use the separate system calls, except for accept() which may need to be a wrapper for accept4(). Remove support for socketcall(), and move the socket system call definitions into SYSCALLS.def. Signed-off-by: Ben Hutchings <ben@decadent.org.uk>	Ben Hutchings		5 months ago
2238	[klibc] syscalls: Remove _KLIBC_SYS_SOCKETCALL config macro [klibc] syscalls: Remove _KLIBC_SYS_SOCKETCALL config macro From Linux 1.3.44 to 2.6.25, the sparc port implemented separate system calls for socket(), connect(), etc. in its SunOS personality. The system call numbers were defined in <asm/unistd.h> header even though they were not implemented for native Linux programs. As a workaround for this, commit 8594ecec842d ("[klibc] sparc, sparc64 use sys_socketcall unconditionally") added a macro to override automatic detection of these system calls and defined that on these architectures. No other architecture ever needed this, and since Linux 4.20 all of those system calls are now implemented. We can therefore remove the use of this macro. Signed-off-by: Ben Hutchings <ben@decadent.org.uk>	Ben Hutchings		5 months ago
2237	[klibc] syscalls: Make more system calls non-optional [klibc] syscalls: Make more system calls non-optional Currently many system calls are marked as optional in SYSCALLS.def, but should be defined on all architectures. Make them non-optional. Signed-off-by: Ben Hutchings <ben@decadent.org.uk>	Ben Hutchings		5 months ago
2236	[klibc] syscalls: Clean up the fork() and clone() definition... [klibc] syscalls: Clean up the fork() and clone() definitions Currently we're defining __clone() on all architectures, but it's only needed where fork() is missing and it's broken on s390 and s390x where the parameter order is incompatible. We're also building a list of new architectures which don't implement fork(), when we could automatically detect that. To fix these problems: - If __NR_fork is defined, generate a fork() system call wrapper, with only sparc and sparc64 as special cases. - Otherwise, generate __clone() as a clone() system call wrapper. Comment why this works despite the architecture differences. Signed-off-by: Ben Hutchings <ben@decadent.org.uk>	Ben Hutchings		5 months ago
2235	[klibc] syscalls: Define mlock() etc. on all architectures [klibc] syscalls: Define mlock() etc. on all architectures Due to an error in editing preprocessor conditionals long ago, the m{,un}lock{,all}() system call wrappers are only generated on architectures where we use mmap() and not mmap2(). Move the #endif to the right place. Signed-off-by: Ben Hutchings <ben@decadent.org.uk>	Ben Hutchings		5 months ago
2234	[klibc] Remove unused no-MMU support [klibc] Remove unused no-MMU support There is some conditional logic to handle no-MMU architectures, but no such architectures are defined. We are now having to override the automatic detection for each new architecture because it only checks for fork() and not clone(). Since this is currently unused and therefore untested, it's probably broken. Remove it. Signed-off-by: Ben Hutchings <ben@decadent.org.uk>	Ben Hutchings		5 months ago
2233	[klibc] parisc: Fix build with Linux 6.10+ [klibc] parisc: Fix build with Linux 6.10+ The parisc version of <asm/signal.h> did not define _NSIG or struct sigset until recently, so we did it in <klibc/archsignal.h>. In Linux 6.10 and on some stable branches these definitions have been added, conflicting with our definitions. So make ours conditional on !defined(_NSIG). Signed-off-by: Ben Hutchings <ben@decadent.org.uk>	Ben Hutchings		5 months ago
2232	[klibc] Delete the ia64 port [klibc] Delete the ia64 port The Linux ia64 port was finally deleted in v6.7, so there's no point in maintaining support in klibc. Signed-off-by: Ben Hutchings <ben@decadent.org.uk>	Ben Hutchings		6 months ago
2231	[klibc] inet: Stricter IPv6 field parsing in inet_pton() [klibc] inet: Stricter IPv6 field parsing in inet_pton() We currently don't range-check the fields of an IPv6 address, so the following strings are wrongly accepted: "10000::" "::10000" Since we currently only support hexadecimal fields, implement the range check by limiting the number of digits to 4. Signed-off-by: Ben Hutchings <ben@decadent.org.uk>	Ben Hutchings		11 months ago
2230	[klibc] inet: Stricter byte parsing in inet_aton() [klibc] inet: Stricter byte parsing in inet_aton() sscanf() skips spaces in some places, and accepts values that are out-of-range for a byte, so the following strings are wrongly accepted: "1.2.3.256" "256.1.2.3" "-1.2.3.4" "1.2.3.-4" " 1.2.3.4" "1. 2.3.4" "1.2.3.4 " "1.2.3.0x4" Implement our own strict conversion of decimal digits to bytes. Signed-off-by: Ben Hutchings <ben@decadent.org.uk>	Ben Hutchings		11 months ago
2229	[klibc] inet: Fix character type test in inet_pton() [klibc] inet: Fix character type test in inet_pton() The argument to ixdigit() and other ctype functions must be a char value converted to unsigned char, or -1; otherwise the results are undefined. In practice, if char is signed we could passs a value < -1 which would result in an out-of-bounds read. Signed-off-by: Ben Hutchings <ben@decadent.org.uk>	Ben Hutchings		11 months ago
2228	[klibc] tests: Add test programs for inet_aton() and inet_pt... [klibc] tests: Add test programs for inet_aton() and inet_pton() The preceding bug fix provides strong evidence that these functions haven't been tested, so add test programs. Signed-off-by: Ben Hutchings <ben@decadent.org.uk>	Ben Hutchings		11 months ago
2227	[klibc] Fix the inet_pton implementation [klibc] Fix the inet_pton implementation The function inet_pton should read from src but it actually reads from dst. Signed-off-by: Donny Xia <xiadong.main@gmail.com> Signed-off-by: Ben Hutchings <ben@decadent.org.uk>	Donny Xia		11 months ago
2226	[klibc] 2.0.13 released, next version is 2.0.14 [klibc] 2.0.13 released, next version is 2.0.14	Ben Hutchings		1 year ago
2225	[klibc] Install command links as appropriate [klibc] Install command links as appropriate We build gzip to support being called as "gunzip" or "zcat", and similarly we build halt to support being called as "poweroff" or "reboot". In the build directory, we create the aliases as links to the same binary. When installing, however, we pass all the names for the same binary to the "install" command, creating multiple copies of it in the install directory. To fix this: - Define an install_links command that creates links as specified in an install-links-y variable. - Invoke install_links after installing binaries. - In usr/gzip/Kbuild and usr/utils/Kbuild, specify the links in install-links-y instead of install-y. Reported-by: Benjamin Drung <benjamin.drung@canonical.com> References: https://bugs.launchpad.net/ubuntu/+source/klibc/+bug/2028571 Signed-off-by: Ben Hutchings <ben@decadent.org.uk>	Ben Hutchings	klibc-2.0.13	1 year ago
2224	[klibc] kinit: Support specifying root with PARTLABEL [klibc] kinit: Support specifying root with PARTLABEL Allow specifying the root device to be mounted as root=PARTLABEL=label. The label is the GPT partition label of the intended root partition, as presented by the kernel with the uevent entry PARTNAME. Signed-off-by: Rob Vandermeulen <rvandermeulen@google.com> Signed-off-by: Ben Hutchings <ben@decadent.org.uk>	Rob Vandermeulen		1 year ago

Older »