If a process goes away while reading its fields, but we already have
that process in the list, we should keep it in case the "highlight dying
processes" mode is active. Not only is that expected in this mode, but
this should also ensure parents are in the list when their children are
(wanted for tree mode consistency).
A process can die between reading the directory listing and opening the
directory FD (if HAVE_OPENAT) or /proc files (otherwise) for reading the
process data. This race would cause LinuxProcessList_recurseProcTree to
remove it from the list immediately, which is unexpected in the
"highlight dying processes" mode and can break the tree structure.
This patch closes this race in the HAVE_OPENAT case by only accessing
the process entry after the directory FD has been opened.
man sscanf(3):
A sequence of white-space characters (space, tab, newline, etc.; see isspace(3)).
This directive matches any amount of white space, including none, in the input.
At the moment this is used to make the memory meter report sane values even
if the host has ZFS and that leaks through into a containerized environment
Fixes#863
Includes a clever check for magic PROC_PID_INIT_INO in /proc/self/ns/pid thanks to Pavel Snajdr (snajpa)
SELinux contexts can be quite long; adjust the column width dynamically
at each cycle to the longest value.
Also with the recent addition of multiple screens, over-long columns can
be moved into their own screen.
Not all batteries entries in /sys/class/power_supply start with either
BAT or AC, but might have device specific names, e.g. CMB1.
Detect the types of those entries and parse them accordingly.
Closes: #881
Fixes: 3e70de64 ("Code clean up for reading battery info")
Do not underflow count at the last iteration, which triggers UBSAN when
using -fsanitize=unsigned-integer-overflow. This is useful as those
underflows can be a result of a flawed counting logic (e.g. a counter
gets reduced more than increased).
Use PF_KTHREAD flag in /proc/[pid]/stat to detect kernel threads.
This fixed an issue when a process's cmdline is empty, htop think
it is a kernel thread.
While most Unix-like systems use 16-bit user IDs,
Linux supports 32-bit UIDs since version 2.6.
UIDs above 65535 are used for UID namespacing of containers,
where a container has its own set of 16-bit user IDs.
Processes in such containers will have (much) larger UIDs than 65535.
Because the current format strings for `ST_UID` and `USER`
are `%5d` and `%9d` respectively, processes with such UIDs
lead to misaligned columns.
Dynamically scale the `ST_UID` column and increase the size of `USER`
to 10 characters (length of UINT32_MAX) to ensure that the user ID always fits.
Additionally: clean up how the titlebuffer size calculation and ensure
the PID column has a minimum size of 5.
Limit the maximum width (instead of only the minimum width), pad the
header width accordingly, and also remove extra stray spaces from the
format string (the main spacing should just come from the alignment of
the value).
Fixes#850.
This is real, physical memory available for applications to
use. We should not try to pretend otherwise; its confusing
for users and inconsistent with all other tools.
Add an explicit else clause so a following else branch for a prior if
condition does not get mixed up.
Also force a trailing semicolon and thereby silence current
-Wextra-semi-stmt warnings.
Improve readability of the hwloc_bitmap_foreach_begin loop macro.
A process, whose executable has been replaced and thus marked by htop,
can be re-executed with the replaced executable, with the same PID, in
two ways: the Linux feature checkpoint/restore or re-execution of PID 1.
The actual check is just a string comparison, like the dropped
condition, leading to (almost) no computation overhead.
Zero all the CPU data, like totalPeriod, after its memory allocation via
realloc(3).
Conditional jump or move depends on uninitialised value(s)
at 0x132A9B: LinuxProcessList_scanCPUTime (LinuxProcessList.c:1928)
by 0x1358C3: ProcessList_goThroughEntries (LinuxProcessList.c:2079)
by 0x12A79A: ProcessList_scan (ProcessList.c:627)
by 0x11CA67: CommandLine_run (CommandLine.c:357)
by 0x4A81E49: (below main) (libc-start.c:314)
Uninitialised value was created by a heap allocation
at 0x48396C5: malloc (in /usr/libexec/valgrind/vgpreload_memcheck-amd64-linux.so)
by 0x12F633: xRealloc (XUtils.c:64)
by 0x12F633: xReallocArray (XUtils.c:78)
by 0x1325A8: LinuxProcessList_updateCPUcount (LinuxProcessList.c:207)
by 0x134E0A: ProcessList_new (LinuxProcessList.c:284)
by 0x11C8D0: CommandLine_run (CommandLine.c:301)
by 0x4A81E49: (below main) (libc-start.c:314)
Generalize sub-diskname handling, like sdb1/sdb2, to not count the
usage twice with the aggregate top-diskname, like sdb.
Rely on /proc/diskstats being ordered, e.g. no sub-diskname precedes its
top-diskname.
Closes: #675
Daniel Lange