That's right. You can write your own implementation and install it like the others (arch/x86/linux/syscalls.cc) if you'd like. This system call is tied to multithreading, so you might have to implement a lot of other things to get it to work correctly. If you know it's being called in a way where it's not important that it actually does anything, you can tell M5 to ignore it. There are examples of that in the same file.
Gabe Sitos Lin wrote: > Hi Gabe: > The followings are the messages when I try to execute a x86 binary > on M5 x86_se: > > Global frequency set at 1000000000000 ticks per second > 0: system.remote_gdb.listener: listening for remote gdb #0 on port 7000 > **** REAL SIMULATION **** > info: Entering event queue @ 0. Starting simulation... > fatal: syscall set_tid_address (#218) unimplemented. > @ cycle 606500 > [unimplementedFunc:build/X86_SE/sim/syscall_emul.cc, line 73] > Memory Usage: 2162020 KBytes > > It seems that the system call "set_tid_address" is not supported in the M5. > > regards, > sitos > > 2009/1/23 <[email protected]>: > >>> 2) I tried to use the newest M5 and run the X86_SE or X86_FS mode >>> with the binaries compiled from conventional x86 gcc compiler. The >>> status is even worse.... >>> >> Could you please let me know what didn't work? X86 support is a work in >> progress >> so there may be major limitations like an unimplemented feature that will >> keep >> it from working, but it might be a simple bug. >> >> Gabe >> >> _______________________________________________ >> m5-users mailing list >> [email protected] >> http://m5sim.org/cgi-bin/mailman/listinfo/m5-users >> >> > _______________________________________________ > m5-users mailing list > [email protected] > http://m5sim.org/cgi-bin/mailman/listinfo/m5-users > _______________________________________________ m5-users mailing list [email protected] http://m5sim.org/cgi-bin/mailman/listinfo/m5-users
