aboutsummaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
...
* Restructure argument setup in hashbangFlavio Cruz2024-09-081-48/+43
| | | | | | | | | | | | | | | We do a few things here: - Move search_path to the scope where it is used to make dependencies more clear. - Have a separate variable to store the file name we eventually need to free and move the free logic to happen in a single place. Both of this allows us to still free the name even if a fault is generated and also avoids a compiler warning as we try to assign a 'const char*' file_name_exec to a 'char *', making it more clear to what exactly we need to free. I also believe 'error' in line 245 was not initialized in case 'file_name_exec' is used and this fixes that too. Message-ID: <20240121210757.1900938-1-flaviocruz@gmail.com>
* x86_64: do not define mach_cpu_subtypes since we don't use itFlavio Cruz2024-09-082-0/+4
| | | | Message-ID: <20240121210757.1900938-4-flaviocruz@gmail.com>
* Fix warnings in fstestsFlavio Cruz2024-09-082-3/+8
| | | | Message-ID: <20240121210757.1900938-3-flaviocruz@gmail.com>
* procfs: remove unused rootdir_symlink_make_nodeFlavio Cruz2024-09-081-8/+0
| | | | | Not needed since b2c97e251bb470e6f967c716081675a96dbde59c Message-ID: <20240121210757.1900938-2-flaviocruz@gmail.com>
* remap, syncfs: correct program version namefree_software@xobnur.uk2024-09-072-2/+2
| | | | | /hurd/remap --version was claiming to be fakeroot Message-ID: <621e135a-f272-43e8-9c10-35804abb1ed5@xobnur.uk>
* libports: Add missing ports_interrupt_server declarationSamuel Thibault2024-09-011-0/+2
|
* startup, settrans: Add missing flush of stderr to get output on waitingSamuel Thibault2024-08-292-0/+2
|
* ext2fs: Ignore translators bits when checking we have no Hurd extensionsSamuel Thibault2024-08-261-1/+1
| | | | We don't actually want to write these anyway.
* ext2fs: Fix masking out translators bitsSamuel Thibault2024-08-261-1/+1
| | | | We need to mask out before shifting the bits.
* ext2fs: mention which inode size is supportedSamuel Thibault2024-08-261-1/+1
|
* Fix line-buffered stderrSamuel Thibault2024-08-258-3/+8
| | | | | mach_open_devstream creates a fully-buffered stream by default. This prevents from seeing various messages.
* libdiskfs: Print error if we fail to open /dev/consoleSamuel Thibault2024-08-251-6/+14
|
* Make sure to also print early-fatal errors on mach consoleSamuel Thibault2024-08-252-4/+11
| | | | In case the user is using a debug kernel, they will get to see these.
* Fix build against libgcrypt >= 1.11.0Samuel Thibault2024-07-233-169/+3
| | | | It now provides a pkg-config file instead of libgcrypt-config
* nfsd: fix -Werror=incompatible-pointer-types issues for x86_64Flavio Cruz2024-07-232-3/+3
| | | | Message-ID: <3pmgbk2myx4phdb6wcso54solmybulrvbytols5md3nri2p2qp@xjmfmpvwjp6t>
* netfs and diskfs: make showtrans work correctly when translators are set on ↵Flavio Cruz2024-07-102-35/+32
| | | | | | | | | top of sylink/device/fifo/socket files. While these are short-circuited translators and simply change st_mode, actual translators can be set on top. We want `showtrans` to show the actual passive translator, not the short-circuited one. Message-ID: <b3ujoovkmz3iagqkmv6lyzgqsv3g5u2wiaqwktlttbalyd7oak@6qeiiupzzmag>
* tcp: select(writefds) don't hang up when a peer close connectionKOSAKI Motohiro2024-07-081-1/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This issue come from ruby language community. Below test program hang up when only run on Linux. % uname -mrsv Linux 2.6.26-2-486 #1 Sat Dec 26 08:37:39 UTC 2009 i686 % ruby -rsocket -ve ' BasicSocket.do_not_reverse_lookup = true serv = TCPServer.open("127.0.0.1", 0) s1 = TCPSocket.open("127.0.0.1", serv.addr[1]) s2 = serv.accept s2.close s1.write("a") rescue p $! s1.write("a") rescue p $! Thread.new { s1.write("a") }.join' ruby 1.9.3dev (2010-07-06 trunk 28554) [i686-linux] #<Errno::EPIPE: Broken pipe> [Hang Here] FreeBSD, Solaris, Mac doesn't. because Ruby's write() method call select() internally. and tcp_poll has a bug. SUS defined 'ready for writing' of select() as following. | A descriptor shall be considered ready for writing when a call to an output | function with O_NONBLOCK clear would not block, whether or not the function | would transfer data successfully. That said, EPIPE situation is clearly one of 'ready for writing'. We don't have read-side issue because tcp_poll() already has read side shutdown care. | if (sk->sk_shutdown & RCV_SHUTDOWN) | mask |= POLLIN | POLLRDNORM | POLLRDHUP; So, Let's insert same logic in write side. - reference url http://blade.nagaokaut.ac.jp/cgi-bin/scat.rb/ruby/ruby-core/31065 http://blade.nagaokaut.ac.jp/cgi-bin/scat.rb/ruby/ruby-core/31068 Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* pflocal: make S_socket_recv return 0 instead of EPIPESamuel Thibault2024-07-081-0/+1
| | | | | Callers are not expecting to get EPIPE, and rather a real EOF, like is done in S_io_read.
* Fix issues with GCC 14.Flavio Cruz2024-06-243-32/+16
| | | | | | | | | * ftpfts/ftpfs: use appropriate function signature for interrupt_check. * pfinet/Makefile: turn off warnings in Linux driver code that are now errors in GCC 14. * pfinet/glue-include/linux/socket.h: remove duplicate functions. These are implemented in linux-src/net/core/iovec.c. Message-ID: <pnx5mho5ipgabehscwu2ogk5l7zfjsxpmjz37fuplgbtohwtjj@ofhr2r5xvqm6>
* Include unistd.h for sleep in pfinet/ethernet.cFlavio Cruz2024-06-241-0/+1
| | | | | Compilation fails otherwise due to -Werror=implicit-function-declaration Message-ID: <uuxfc5gjinlekgdyjzxesz72t6dabbl6s2f2n2fi7iw2bi5nl5@js7sx73kagmd>
* pfinet: wait a bit before retrying to open ethernet deviceSamuel Thibault2024-05-191-1/+2
| | | | | If netdde is restarting, it may have not finished probing devices, give it some time.
* startup: wait for any kind of bootstrap filesystemfirst last2024-05-091-1/+1
| | | | | The bootstrap filesystem doesn't have to be ext2fs Message-ID: <8932f426-8ab5-4344-81a6-52a90dfada09@runbox.com>
* isofs: fix rock-ridge detectionfree_software@xobnur.uk2024-05-061-1/+1
| | | | | The field length check should check for field overflow. Message-ID: <8e69aefc-8c08-4189-a5b9-974303427597@xobnur.uk>
* libirqhelp: Add libraryDamien Zammit2024-03-264-0/+436
| | | | | | Add a helper library for attaching interrupt handlers in userspace. Message-ID: <20240326045846.1661099-2-damien@zamaudio.com>
* Revert "exec: Stop relying on address space size"Samuel Thibault2024-03-231-1/+3
| | | | | | This reverts commit db46ea2eb9dc84959fbf9b1819facac3d6078ba1. This is making the hurd startup hang.
* exec: Add support for AArch64 executablesSergey Bugaev2024-03-231-0/+5
| | | | | | | | | This maps to EM_AARCH64. Just like the x86_64 branch, we only compile this code if CPU_TYPE_ARM64 is defined in Mach headers, to avoid raising Mach version requirement on other architectures; and we explicitly enable the branch when building for AArch64 itself, to get a build error if CPU_TYPE_ARM64 is somehow not defined. Message-ID: <20240323115322.69075-7-bugaevc@gmail.com>
* boot: Add support for AArch64Sergey Bugaev2024-03-231-0/+11
| | | | Message-ID: <20240323115322.69075-10-bugaevc@gmail.com>
* proc: Add support for AArch64 in unameSergey Bugaev2024-03-232-2/+5
| | | | | | | | | | | | | Since no CPU subtypes are defined for CPU_TYPE_ARM64, just report the type, same as for x86_64. Sample uname(2) output: sysname: GNU release: 0.9 version: GNU-Mach 1.8/Hurd-0.9 machine: aarch64 Message-ID: <20240323115322.69075-9-bugaevc@gmail.com>
* elfcore: Add support for saving AArch64 registersSergey Bugaev2024-03-231-0/+26
| | | | Message-ID: <20240323115322.69075-8-bugaevc@gmail.com>
* exec: Stop relying on address space sizeSergey Bugaev2024-03-231-3/+1
| | | | | | | The code here just wants to deallocate the whole address space, and Mach already contains the logic to limit the passed-in range to the vm_map's actual bounds (see VM_MAP_RANGE_CHECK). Message-ID: <20240323115322.69075-6-bugaevc@gmail.com>
* libshouldbeinlibc: Stop relying on address space sizeSergey Bugaev2024-03-231-1/+1
| | | | | | | | | | While GNU Mach on AArch64 still exports VM_MIN_ADDRESS / VM_MAX_ADDRESS for compatibility, we should try to rely on it less when possible; in the future we might be able to stop exporting them from Mach. The code here really just wants to wire everything in its address space, and the wire_segment_internal () routine already queries for actually present memory regions dynamically. Message-ID: <20240323115322.69075-5-bugaevc@gmail.com>
* proc: Only try host_kernel_version () on i386Sergey Bugaev2024-03-231-2/+2
| | | | | | None of the new architectures are going to have it; that isn't specific to x86_64. Message-ID: <20240323115322.69075-4-bugaevc@gmail.com>
* Make long & friends 64-bit on 64-bit platformsSergey Bugaev2024-03-231-1/+1
| | | | | Not only on x86_64. Message-ID: <20240323115322.69075-3-bugaevc@gmail.com>
* exec: Fix creating executable stacksSergey Bugaev2024-03-231-4/+5
| | | | | | | | | | | | | | | | The previous logic had two independent issues: * We need to make the stack executable if either the program or its ELF interpreter requires executable stack. In practice, it's common for the program itself to not require executable stack, but ld.so (glibc) needs it. * mach_setup_thread () allocates stacks with a simple vm_allocate (), which creates non-executable memory. So if an executable stack is required, the stack has to be vm_protect'ed to enable execution, not the other way around. As the comment suggest, it would've been better to use vm_map () with the desired protection directly. Message-ID: <20240323115322.69075-2-bugaevc@gmail.com>
* term: Fix function prototypeSergey Bugaev2024-03-231-1/+1
| | | | | | | | struct bottomhalf.mdmstate is of type error_t (*)(int *state). Fixes -Wincompatible-pointer-types, which is a hard error by default in GCC 14. Message-ID: <20240323115322.69075-1-bugaevc@gmail.com>
* console-client: Add another missing error checkEtienne Brateau2024-03-121-0/+2
| | | | Message-ID: <20240312220842.79072-1-etienne.brateau@gmail.com>
* console-check: Add missing error checkEtienne Brateau2024-03-121-0/+2
| | | | Message-ID: <20240312220311.76709-1-etienne.brateau@gmail.com>
* ext2fs: support translators on non-hurd-created filesystemsSamuel Thibault2024-03-121-36/+44
| | | | | We are fine with only using xattr on filesystems that don't have the i_translator inode field.
* console-client: use xkbcommon instead of x11 for xkb extended supportEtienne Brateau2024-03-1019-5768/+452
| | | | | | | | | | | | | | | | | | This allow to reduce the dependencies, only xkbcommon (keyboard support only) is required instead of the whole x11 library + lex + yacc. This replacement allow to reduce the code size, now features are handled by xkbcommon itself. The functionnalites remain the sames (actions are reimplemented but in the code directly as it’s impossible to add custom actions). The custom xkb data files are removed as we can now directly use the standard ones from xkeyboard-config. The configuration to launch the console keyboard modules changed to now directly configure the model+layout+variat+options directly. Tested by compiling with and without xkbcommon. Tested X11 (ran i3 correctly). Composing is still working. Message-ID: <20240309234838.31923-1-etienne.brateau@gmail.com>
* console-client: use constant value instead of magic numberEtienne Brateau2024-03-091-1/+1
| | | | Message-ID: <20240309003840.586490-1-etienne.brateau@gmail.com>
* MAKEDEV: when creating devices, ensure the underlying files are either ↵Flavio Cruz2024-03-081-25/+39
| | | | | | | | | | | | | block/char devices or directories The perl test suite has a test where it reads all the block or char devices under /dev without following the translators. Then it compares it against a list of devices that read the translated nodes stat info. The patch changes how the the device files are created initially so that the stat information is identical and makes the Hurd environment appear more similar to other operating systems. Message-ID: <Zef__LrgPLXp9WHG@jupiter.tail36e24.ts.net>
* procfs: fix MemTotal 4G memory clamp on 32bit systemsSamuel Thibault2024-03-061-1/+4
| | | | Using vmstats allows to get up to 16T.
* ext2fs: New default: use xattrs to store translator recordsDamien Zammit2024-03-051-9/+9
| | | | | | | | | | Replaces experimental option --x-xattr-translator-records with --no-xattr-translator-records to allow rolling back to previous behaviour. NB: - Legacy records still work with either setting. - Adding a new record removes a legacy one.
* ext2fs: Check that the filesystem supports xattr before reading itDamien Zammit2024-03-041-10/+15
|
* libps: update ps_emit_nice_size_t to handle arbitrarily large size_tFlavio Cruz2024-03-031-9/+13
| | | | | | | Update argument types for sprint_frac_value to reflect how big they actually are so that GCC doesn't think it needs a larger buffer than necessary. Message-ID: <ZeS1i5u_OziWpApt@jupiter.tail36e24.ts.net>
* libbpf: avoid aborting on unknown instructionSamuel Thibault2024-03-011-1/+2
| | | | | | | Userland might load BPF programs with unknown instructions, we currently don't pre-check against that. In such a case, we shouldn't make netdde completely abort, and rather just return 0 like e.g. in the division by zero case.
* rumpdisk: do not open device if block size is 0Flavio Cruz2024-03-011-6/+17
| | | | | | | | | | | | | | | | | Currently, if we do: $ ls /dev/cd0/ The computer seems to get stuck, caused by the divide by 0 in the rumpdisk server in device_get_status. I noticed that if we have no disk in the cdrom device, we can still open it but block and media size will be 0 and the message "cd0 dos partition I/O error" will be printed to the console. To avoid this problem, we check the block size and throw an error when it is 0. This also works correctly when a disk actually exists. This should help fix the perl and likely the vim test suites that are currently failing in https://buildd.debian.org/. Message-ID: <Zd_8XjcHcbNIp5NM@mars.tail36e24.ts.net>
* Add proc_getchildren_rusage RPC and track rusage for children and descendantsFlavio Cruz2024-02-174-1/+19
| | | | Message-ID: <20240216182630.5770-2-flaviocruz@gmail.com>
* runsystem.hurd.sh: Refresh from runsystem.hurd.sh.inSamuel Thibault2024-02-081-10/+10
|
* runsystem: Check that /servers/socket/1 is really set upSamuel Thibault2024-02-082-2/+2
| | | | | | In case the image was built through a tarball, /servers/socket/1 might exist but not actually have been configured as pflocal translator. So better check that we do have a translator there, and fix it otherwise.
generated by cgit v1.2.3 (git 2.39.1) at 2025-04-04 11:00:56 +0000