SamyGO

patois wrote:I'm currently stuck at trying memjack to compile (I have stated this in another thread but I think this thread should be more appropriate).

Code: Select all

dennis@ubuntu:~/Downloads/memjack$ make
arm-v7a8v2r2-linux-gnueabi-gcc memjack.o util.o procutils.o -o memjack
arm-v7a8v2r2-linux-gnueabi-gcc: selected multilib '.' not installed
make: *** [memjack] Error 1

Does anyone have an idea what the problem is and how to fix it?

maybe some env problem?
http://comments.gmane.org/gmane.comp.gcc.help/39888

no idea if its related to toolchain from samygo svn. im using toolchain from samsung.

Hmm.. found this via google as well. I think I'll just try the samsung toolchain. Thanks!

edit: the samsung toolchain worked flawlessly, thanks!

Topic splitted.
viewtopic.php?f=27&t=6325

VDLinux#> ./hijackx -p `pidof exeTV` -l /mtd_rwarea/test/dumpkey.so
VDLinux#> cat dumpkey.log
file: /dtv/usb/sda2/CONTENTS/20130820124418.srf
key: {0x0d, 0xaa, 0xf1, 0x49, 0x2b, 0xe1, 0xe2, 0xb9, 0x42, 0x4d, 0x75, 0x4b, 0xcf, 0x9e, 0x9a, 0x18, }

channelinfo patch for f-series
add this to mem_patches_exeAPP in memjack

new version of hijack, see first post.

Look, this way for each so file we need to execute hijack once. Why don`t you add to patcher usage of arguments? i mean $1 $2 $3 .... and if you have a lot of so files to inject, it is enough to execute injector just once. or do you think this is obsolete?

i think a better way would be to inject a "loader .so" that e.g. reads a config and loads all the .so to be loaded.
also there might be different requirements. .so that hook stuff need to be permanent in memory while .so files that just patch some stuff dont.

Okey, loader is ok too. more robust, i agree.

first post updated