On some systems like Chromium OS, loading modules from non-verified
filesystems is denied. Only finit_module is allowed because an open
fd is passed which can be checked against a verified location.
Change the module loading code to first attempt finit_module and if
that fails for whatever reason, fall back to the existing logic.
On x86_64, this adds ~80 bytes to modutils/modutils.o and ~68 bytes
to modutils/modprobe-small.o.
Signed-off-by: Mike Frysinger <vapier@chromium.org>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
"modprobe minix; echo $?"
Was:
modprobe: corrupted data
modprobe: read error from 'kernel/fs/minix/minix.ko.xz': No such file or directory
modprobe: corrupted data
modprobe: read error from 'kernel/fs/minix/minix.ko.xz': No such file or directory
modprobe: corrupted data
modprobe: read error from 'kernel/fs/minix/minix.ko.xz'
modprobe: 'kernel/fs/minix/minix.ko.xz': Success
0
Now:
modprobe: corrupted data
modprobe: read error from 'kernel/fs/minix/minix.ko.xz'
1
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
This matches behavior with kmod which has been the standard for a long
time at this point.
URL: https://bugs.busybox.net/8021
Reported-by: Jö <jorrit@jorrit.de>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
usecase:
two sd cards are being mounted in parallel at same time on dual core. example
modules which are getting loaded is nls_cp437. While one module is being
loaded , it makes state in /proc/modules as 'coming' and then starts doing its
module init function (in our case - registering nls). meanwhile on other core,
if modprobe returns that is has already been loaded, then it will continue
and search for the nls list which is not yet finished from first module init.
This fails resulting in not mounting sd card.
function old new delta
process_module 667 746 +79
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
"rmmod OUT_OF_TREE_MODULE" was not working, because module is not in depmod file.
In general, rmmod doesn't need scanning, it simply unloads every argv[i].
function old new delta
rmmod - 63 +63
modprobe_main 449 465 +16
process_module 705 667 -38
------------------------------------------------------------------------------
(add/remove: 1/0 grow/shrink: 1/1 up/down: 79/-38) Total: 41 bytes
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
This fixes a regression introduced with commit 78854520 (modprobe:
revert checking for /, stop doing basename() on modprobe args,
2015-01-01) that broke modprobe options stored in /etc/modprobe.conf and
/etc/modprobe.d/*.conf.
This also fixes modprobe -r
Signed-off-by: Natanael Copa <ncopa@alpinelinux.org>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
"modprobe -r MODNAME", after it found the full module pathname,
should strip dirpath before trying to remove the module.
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
display() function returns the length of the value we are displaying + 1.
As a consequence, if we have field=value\0field=value\0field=value in the
binary, then the second occurence will be skipped as ptr will miss the first
character of the following field.
Example trying to list aliases from ixgbe.ko on a 3.2 kernel.
- In the module we have:
alias=pci:v00008086d00001560sv*sd*bc*sc*i*\0
alias=pci:v00008086d0000154Asv*sd*bc*sc*i*\0
alias=pci:v00008086d00001557sv*sd*bc*sc*i*\0
alias=pci:v00008086d0000154Fsv*sd*bc*sc*i*\0
alias=pci:v00008086d0000154Dsv*sd*bc*sc*i*\0
...
- Using modinfo -F alias ixgbe returns:
alias: pci:v00008086d00001560sv*sd*bc*sc*i*
alias: pci:v00008086d00001557sv*sd*bc*sc*i*
alias: pci:v00008086d0000154Dsv*sd*bc*sc*i*
...
This problem appeared with kernel commit "modules: no need to align .modinfo
strings" b6472776816af1ed52848c93d26e3edb3b17adab in 2.6.37.
Fix this by not trusting display() return value but increment ptr by strlen(ptr)
(the same way as depmod.c does).
Signed-off-by: David Marchand <david.marchand@6wind.com>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
Closes 627 and 7034.
Commonly seen case is (un)loading of an alias
which matches ata_generic and a more specific ata module.
For example:
modprobe [-r] pci:v00008086d00007010sv00000000sd00000000bc01sc01i80
(ata_generic and pata_acpi)
modprobe [-r] pci:v00001106d00000571sv00001509sd00009022bc01sc01i8a
(ata_generic and pata_via)
function old new delta
process_module 615 728 +113
parse_module 309 395 +86
find_alias 621 653 +32
pathname_matches_modname 78 79 +1
------------------------------------------------------------------------------
(add/remove: 0/0 grow/shrink: 4/0 up/down: 232/0) Total: 232 bytes
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
The modern module-init-tools 3.12 has "-D" option and it works.
The --show-depends and -a options can be used together due to upstream modprobe
behaviour.
function old new delta
do_modprobe 536 579 +43
packed_usage 28035 28042 +7
modprobe_main 653 657 +4
------------------------------------------------------------------------------
(add/remove: 0/0 grow/shrink: 3/0 up/down: 54/0) Total: 54 bytes
Signed-off-by: Serj Kalichev <serj.kalichev@gmail.com>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
If there are some spaces in the insmod command line, then
this will be splitted in single words as separate elements of
argv. It just needs to chain them together in the options string
passed to the sys_init_module syscall.
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com>
This change retains "or later" state! No licensing _changes_ here,
only form is adjusted (article, space between "GPL" and "v2" and so on).
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
Add support of NDS32 architecture to modutils.
Signed-off-by: Macpaul Lin <macpaul@andestech.com>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
PLATFORM_LINUX is used as a dependency for applets or features
which require Linux-specific interfaces.
Signed-off-by: Jeremie Koenig <jk@jk.fr.eu.org>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
Before: modprobe -l
modules.dep not found
No size change, just moved the chdir.
Signed-off-by: Lauri Kasanen <curaga@operamail.com>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
function old new delta
modinfo 272 329 +57
modinfo_main 325 344 +19
Signed-off-by: Lauri Kasanen <curaga@operamail.com>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
Fixed some goofs in help text
function old new delta
modprobe_main 494 653 +159
packed_usage 27129 27145 +16
insmod_main 98 95 -3
Signed-off-by: Pascal Bellard <pascal.bellard@ads-lu.com>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
Based on patch by Ozan Çağlayan (ozan AT pardus.org.tr)
function old new delta
parse_and_add_kcmdline_module_options - 149 +149
do_modprobe 357 365 +8
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
Also opens up a possibility to make other unicode stuff smaller
and more correct later. but:
function old new delta
static.combining - 516 +516
bb_wcwidth - 328 +328
unicode_cut_nchars - 141 +141
mbstowc_internal - 93 +93
in_table - 78 +78
cal_main 899 961 +62
static.combining0x10000 - 40 +40
unicode_strlen - 31 +31
bb_mbstrlen 31 - -31
bb_mbstowcs 173 102 -71
------------------------------------------------------------------------------
(add/remove: 7/1 grow/shrink: 1/1 up/down: 1289/-102) Total: 1187 bytes
Uses code of Markus Kuhn, which is in public domain:
http://www.cl.cam.ac.uk/~mgk25/ucs/wcwidth.c
"Permission to use, copy, modify, and distribute this software
for any purpose and without fee is hereby granted. The author
disclaims all warranties with regard to this software."
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
Added O_RDONLY where improves readability. Note: O_RDONLY == 0,
so it is there even if not specified.
function old new delta
stty_main 1289 1235 -54
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
