According to git blame the trymmap() function was added in 1996 to skip
mmap() calls for NFS file systems. However, nowadays mmap() should be
perfectly safe even on NFS. Importantly, onl ufs and cd9660 file systems
were whitelisted so we don't use mmap() on ZFS. It also prevents the use
of mmap() when bootstrapping from macOS/Linux since on those systems the
trymmap() function was always returning zero due to the missing MFSNAMELEN
define.
This change keeps the trymmap() function but changes it to check whether
using mmap() can reduce the number of system calls that are required.
Using mmap() only reduces the number of system calls if we need multiple read()
syscalls, i.e. if the file size is > MAXBSIZE. However, mmap() is more expensive
than read() so this sets the threshold at 4 fewer syscalls. Additionally, for
larger file size mmap() can significantly increase the number of page faults,
so avoid it in that case.
It's unclear whether using mmap() is ever faster than a read with an appropriate
buffer size, but this change at least removes two unnecessary system calls
for every file that is installed.
Reviewed By: markj
Differential Revision: https://reviews.freebsd.org/D26041
GLibc expects six 'X' characters in the mkstemp template argument and
will return EINVAL otherwise.
Reviewed By: emaste, imp, mjg
Differential Revision: https://reviews.freebsd.org/D25662
Make sure we call fsync(2) on strip result
in case of "safecopy" and "strip -o tempcopy -- src"
before renaming tempcopy to destination.
MFC after: 3 weeks
X-MFC-With: r363064
Also, make it not break if STRIPBIN points to strip version without -o support
and destination does not exist before installing.
Reported by: lwhsu
MFC after: 1 month
X-MFC-With: 363064
Make it not break if STRIPBIN points to strip version without -o support.
In that case, perform extra copy just like before r363064.
MFC after: 1 month
X-MFC-With: 363064
Currently, "install -s -S" behaviour is inefficient for upgrade.
First it finds that destination file already exists and copies
source file to temporary file. Then it calls strip(1)
with name of temporary file as single agrument and our strip(1) creates
another temporary file in the /tmp (or TMPDIR) making another copy
that is finally copied to DESTDIR third time.
Meantime, strip(1) has an option "-o dst" to specify destination
so install(1) is allowed to skip initial copying from obj to DESTDIR.
This change makes it do so.
Take a look at https://reviews.freebsd.org/D25551 for details
and efficiency numbers (in short: upto 32% gained for installword).
MFC after: 1 month
Differential Revision: https://reviews.freebsd.org/D25551
While here:
- Remove deprecated Tn macros.
- Do not use macros with the -width option.
Reviewed by: bcr
Approved by: doc (bcr)
Differential Revision: https://reviews.freebsd.org/D21173
For example, from the referenced PR [1]:
$ mkdir /tmp/lib/ /tmp/libexec
$ touch /tmp/lib/foo.so
$ install -lrs /tmp/lib/foo.so /tmp/libexec/
The common path identification bits terminate src at /tmp/lib/ and the
destination at /tmp/libe. The subsequent backtracking is then incorrect, as
it traverses the destination and backtraces exactly one level while eating
the 'libexec' because it was previously (falsely) identified as common with
'lib'.
The obvious fix would be to make sure we've actually terminated just after
directory separators and rewind a character if we haven't. In the above
example, we would end up rewinding to /tmp/ and subsequently doing the right
thing.
Test case added.
PR: 235330 [1]
MFC after: 1 week
I need these tools in order to install the crossbuilt FreeBSD and create a
disk image. Linux does not have a st_flags in struct stat so unfortunately
I need a bunch of ugly ifdefs. The resulting binaries allow me to
sucessfully install a MIPS64 world and create a disk-image that boots.
Reviewed By: brooks, bdrewery, emaste
Approved By: jhb (mentor)
Differential Revision: https://reviews.freebsd.org/D13307
Mainly focus on files that use BSD 3-Clause license.
The Software Package Data Exchange (SPDX) group provides a specification
to make it easier for automated tools to detect and summarize well known
opensource licenses. We are gradually adopting the specification, noting
that the tags are considered only advisory and do not, in any way,
superceed or replace the license texts.
Special thanks to Wind River for providing access to "The Duke of
Highlander" tool: an older (2014) run over FreeBSD tree was useful as a
starting point.
If stat continues to fail with ENOENT and mkdir with EEXIST the code wont
finish. In particular this can show up when the target path follows through
a symlink to a non-existent directory.
Reported by: ae
MFC after: 1 week
Prefer ${SRCTOP}/foo over ${.CURDIR}/../../foo and ${SRCTOP}/usr.bin/foo
over ${.CURDIR}/../foo for paths in Makefiles.
Differential Revision: https://reviews.freebsd.org/D9932
Sponsored by: Netflix
Silence on: arch@ (twice)
The additional testcases use absolute paths for sources and targets,
as the other testcase which tested `-l sr` used flat relative paths in
the same directory.
Please note that these testcases do not test `-l a` -- that's already
addressed in the battery of tests.
MFC after: 1 week
Sponsored by: Dell EMC Isilon
Unfortunately kyua does not omit the path mismatch on failure, so it must be coded
into the error message.
Cache the values, run the test(1) call, then print out the values in an atf_fail
call to emit the required diagnostics to debug why things are failing.
MFC after: 1 week
Sponsored by: Dell EMC Isilon
It was unlinking the target even though it uses rename(2) which already
effectively does that. -S is intended to not unlink(2) the target first.
MFC after: 1 week
Reviewed by: jhb
Sponsored by: EMC / Isilon Storage Division
Differential Revision: https://reviews.freebsd.org/D7452
With both of these flags, the backup was created via rename(dest, backup)
followed by rename(tmp, dest). This left the destination file missing
for a moment which contradicts the point of -S.
This fixes a race with installworld where PRECIOUSPROG and PRECIOUSLIB
files (which use -S for installation) would briefly be missing. In the
case of installing rtld with parallel installworld it could render an
error due to not having rtld present to run install/cp in another
process.
Reported by: jhb
Reviewed by: jhb
MFC after: 1 week
Sponsored by: EMC / Isilon Storage Division
Differential Revision: https://reviews.freebsd.org/D7451
This is a minor nit after r289391 made all installations to a directory always
end in a trailing '/'.
MFC after: 3 days
Sponsored by: EMC / Isilon Storage Division
This should fix the build on older stable/10, since install is a bootstrap
tool.
Pending a decision how to fix this properly, revert utimensat usage. Copies
with the -p option will again appear older than the original almost always,
but -p is not commonly used.
The main problem was bitrot after elftoolchain being swapped in for the
GNU toolchain.
This also reworks how the list of 'host allowed' libraries is determined
to only allow INTERNALLIBs, which is needed for libelftc to come in.
For usr.bin/readelf use the same hack, as libelf and libdward, to bring in
the needed sys/ headers for host builds. This has not yet been a problem due
to readelf not being built as a host tool in buildworld. This is possible
in the meta build though when building the toolchain.
Sponsored by: EMC / Isilon Storage Division
In case the target of install is a dead symlink, install(1) used to not
consider it as "existing" because of the usage of stat(2) instead of
lstat(2). meaning the old file (the symlink) is not removed before the new
file is created. The symlink is being followed and the new file becoming the
target of the symlink instead of the target of install(1)
Reviewed by: jhb, brooks
MFC after: 2 weeks
Differential Revision: https://reviews.freebsd.org/D4191
The bootstrap-tools are supposed to be host tools, which in most cases, use
host headers and libraries. As such, directly including the src tree's headers
for libmd here causes the need to link libmd in since it will be built with
the new symbols (which /usr/lib/libmd.so) won't have unless it is new enough.
During the target build in buildworld the target headers are staged into
WORLDTMP and used via --sysroot, allowing the target xinstall to be built with
the new/target libmd.
The .PATH here was also not doing anything since xinstall does not use libmd
source files.
Sponsored by: EMC / Isilon Storage Division
MFC after: 2 weeks
When requesting install(1) to only make relative symlinks, by pass all the
done to actually compute the relative symlink if the path given in argument is
already a relative path
Off by default, build behaves normally.
WITH_META_MODE we get auto objdir creation, the ability to
start build from anywhere in the tree.
Still need to add real targets under targets/ to build packages.
Differential Revision: D2796
Reviewed by: brooks imp
Since the code stats and mkdirs in 2 separate steps, it is possible that
the directory will be created in the meantime by something else (e.g.
concurrent install).[1]
While here alter the code to properly report stat failure, previously it
would always claim it was mkdir which failed.
Noted by: royger [1]
MFC after: 1 week
