| Age | Commit message (Collapse) | Author |
|---|
|
COLOR_PAIR takes a color pair, it does not produce one.
|
|
that pass through to /dev/usbdisk.
|
|
* libpager/demuxer.c(pager_start_workers): set current and max RLIMIT_AS
to RLIM_INFINITY when the current user has access to the privileged host
port.
Message-ID: <20250616225815.844-2-dnietoc@gmail.com>
|
|
Fall back to bios defaults for requests for irq 9.
Ideally we could check the PIIX3 bridge device exists on
pci, but that would require pci access before pci-arbiter
exists. This is a convenient workaround for now.
Message-ID: <20250622073557.188295-1-damien@zamaudio.com>
|
|
|
|
|
|
|
|
This was offsetting all values on the parameters stack
|
|
Previously, ‘accept’ would return an O_NONBLOCK socket if the listening
socket was O_NONBLOCK at the time the connection was made. With this
change, ‘accept’ always returns a socket where O_NONBLOCK is cleared.
|
|
|
|
Use 64bit mapped time values in maptime_read when the kernel and the
mapped_time_value structure in header file time_value.h support it.
Otherwise step back to use the 32bit time.
Message-ID: <20250503132808.15359-1-zhmingluo@163.com>
|
|
54c0b9b9dbf7 ("ext2fs: Trap trying to access bogus data areas") added
checks on the allocated block number, but did not take care of the
out-of-space condition, which callers of ext2_new_block and
ext2_alloc_block know to handle.
|
|
This fixes making diskfs_catch_exception recursive, catching the last
fault exception, and fixes callers.
|
|
tmpfs loses files when they are unlinked but still mapped.
|
|
Message-ID: <20250402092354.293338-1-zhmingluo@163.com>
|
|
So that if we get a lazily-cleaned inode, we do not get disturbed by the
values, notably i_file_acl and i_dir_acl were not cleaned, and
i_block_group not checked.
b0ff48880bb40 ("Formerly ialloc.c.~6~") introduced only setting
dn_set_ctime when cleanup was missing, but we always want to set the
ctime for a new inode anyway.
|
|
Since it can free more than one block.
|
|
Inline data in i_data is only used by symlinks (and apparently some
device nodes in linux). For regular files and directories we don't store
data there. This is actually important since otherwise
int fd = open("foo.txt", O_WRONLY|O_CREAT);
ftruncate(fd, 1024);
ftruncate(fd, 10);
leads to trying to frob beyond i_data end.
|
|
i.e. superblock, block group descriptor table or beyond the end.
|
|
initscripts apparently tries to mount /run before calling swapon
|
|
i.e. superblock, block group descriptor table or beyond the end.
|
|
This reverts commit 86ea895b1a208925dfc034fac22db2a512a3fab8.
We are really not supposed to have failed to find the block here, it
should have been allocated by pager_unlock_page already.
|
|
find_block ignores EINVAL errors, we need to reconstruct it
|
|
|
|
They really should not be split for the application, and really should
not be queued several times because otherwise we would record the source
several times and get a reference miscount.
|
|
|
|
It may have changed while sleeping.
|
|
While sleeping in pthread_hurd_cond_timedwait_np, current->signal will be
overwritten by the management of other RPCs, so we have to give it some
value on wake up.
Also, if we previously got interrupted, we shouldn't ever try to wait again
until exiting from the RPC: an interrupt means we really want to try hard to
finish the RPC.
Thanks Zhaoming Luo <zhmingluo@163.com> for the deep investigation and
draft!
|
|
For >4G size we need to force 64bit computation on 32bit systems.
|
|
E.g in the "Negative lookup cached." error case, diskfs_lookup clears
NP, so we would segfault.
|
|
* libshouldbeinlibc/lcm.c: make gcd static since it's not exposed as a
symbol.
* pfinet/linux-src/net/ipv4/{tcp,udp}_ipv4.c: Drop inline from lookup functions
since they are used in another module (icmp.c) and shouldn't be removed.
* term/munge.c: make poutput static since it's not exposed as a symbol.
Message-ID: <nmok54owdbglalqkw2ky76jzvn4vvyybpqtftruzhpixcsx7fo@jbphrribcxq5>
|
|
Which allows not to start it unless swap is configured, so Mach can know
when it is not and avoid trying to use a default memory manager.
|
|
It's "supported" only there for now.
|
|
Newer ports won't have it either.
|
|
|
|
|
|
The term_getctty() should accept pty_class. The vim testsuite expects
the master side to say isatty().
See
https://mail.gnu.org/archive/html/bug-hurd/2025-02/msg00061.html
* term/users.c: The term_getctty() accepts pty_class
Message-ID: <20250208103220.853-1-zhmingluo@163.com>
|
|
So the kernel gives us vm_privilege so we can work even when memory is
getting low.
|
|
Instead of killing the whole object.
|
|
|
|
To give explanation to processes suddenly crashing with SIGBUS.
|
|
This avoids
Assertion `!diskfs_readonly' failed.
messages at shutdown.
|
|
|
|
|
|
Next rumpkernel upload will use it.
|
|
|
|
* procfs/rootdir.c: (rootdir_gc_cpuinfo) new function
(rootdir_entries) add entry for cpuinfo file
(cpuinfo_x86, cpuinfo_aarch64) implementations for x86 and
aarch64 respectively.
|
|
when processes which were upgraded didn't get restarted, and stop after
making the filesystem readonly.
|
|
|
|
The usb stack also uses SCSI emulation for usb mass storage.
Message-ID: <20250111082129.1566079-1-damien@zamaudio.com>
|
