[OpenAFS-devel] LWP--stackpointer malloc() being weird on System/390

[Adam Thornton]

Well, I tried reserveFromStack, and that at least reserved some memory
just under 0x80000000, which is more what you'd expect if it were
getting memory from the process stack.  But it's still not remotely
close to what "topstack" is.

So, alternatively, why isn't "topstack" actually something from the
process stack?

Here's the trace:


Adam
savecontext
(0) 0x0   (1) 0x7ffff2c8   (2) 0x8   (3) 0x42fc04   (4) 0x401d18   
(5) 0x1   (6) 0xc010f3c8   (7) 0x80420af0   (8) 0x80420b72   (9) 0x7fffefa8   
***LWP (0x437808): Entered LWP_InitializeProcessSupport
***LWP (0x437808): Entered LWP_CreateProcess
Using reserveFromStack rather than malloc...
Stacksize: 0x30000      Max_stack: 0x8000***LWP (0x437808): Initialized Stackpointer...
Stack pointer after malloc: 0x7ffc70e0
***LWP (0x437808): Entered Initialize_Stack
After Initialize_Stack Stackpointer: 0x7ffc70e0 Stacksize: 0x30000
***LWP (0x437808): Entered Initialize_PCB
savecontext
(0) 0x437180   (1) 0x4379a0   (2) 0x437808   (3) 0x7fff70e0   (4) 0x401d18   
(5) 0x1   (6) 0xc010f3c8   (7) 0x80420af0   (8) 0x80420b72   (9) 0x7ffff080   
***LWP (0x4379a0): Entered Create_Process_Part2
savecontext
(0) 0x437180   (1) 0x4379a0   (2) 0x0   (3) 0x437868   (4) 0x401d18   
(5) 0x1   (6) 0xc010f3c8   (7) 0x80420af0   (8) 0x80420b72   (9) 0x7fffefb8   
stackcheck = 66051: stack = 66051 
topstack = 0x401d18: stackptr = 0x7ffc70e0: stacksize = 0x30000
Wed Jan  3 13:07:53 2001 LWP: stack overflow in process IO MANAGER!
Aborted