Subversion Repositories HelenOS

Rev

Go to most recent revision | Show changed files | Details | Compare with Previous | Blame | RSS feed

Filtering Options

Rev Age Author Path Log message Diff
3661 5709 d 6 h vana /trunk/ IA64: repair similators support  
3659 5709 d 15 h vana /trunk/kernel/arch/ia64/src/ Fix changed param order to ns16550_init function call  
3657 5709 d 15 h vana /trunk/ Legacy IRQ support, uspace NS16550 support, some minor changes  
3655 5710 d 7 h jermar /trunk/kernel/ Add additional members to the irq_t structure so that an interrupt-driven driver
does not need to know how to clear the level interrupt. The z8530 was modified
in this way and is much more generic. The ns16550 driver has also been modified,
but awaits testing. The sparc64 interrupt mapping and dispatch code is now using
the new info and calls the clear-interrupt-routine itself.
 
3654 5710 d 10 h jermar /trunk/kernel/arch/sparc64/ After my effort yesterday, there are two rival psycho drivers.
Merge the two and remove the newer one.
 
3649 5711 d 5 h jermar /trunk/kernel/arch/sparc64/src/drivers/ More work on the psycho driver.  
3648 5711 d 5 h jermar /trunk/kernel/generic/include/ Add host2uintX_t_{le|be} macros to the kernel's byteorder.  
3647 5711 d 7 h jermar /trunk/kernel/arch/sparc64/ Improvized Psycho driver.  
3646 5711 d 8 h jermar /trunk/kernel/ Initial support for interrupt driven driver for ns16550.  
3635 5714 d 7 h vana /trunk/ IA64: Userspace I/O support  
3631 5715 d 6 h jermar /trunk/kernel/arch/ia32/src/ The IA-32 manual is very secretive about the fact that the SYSENTER instruction
disables interrupts (I found only one evidence of this in the instruction's
pseudo code). The sysenter_handler needs to reenable the interrupts in order to
preserve the preemptive nature of the kernel.
 
3626 5716 d 12 h vana /trunk/kernel/ Classic TLB shootdown support on ia64  
3622 5716 d 17 h svoboda /trunk/kernel/generic/include/udebug/ Remove unused debugging task state.  
3620 5716 d 17 h svoboda /trunk/kernel/generic/ Move stuff related to kbox to a separate struct.  
3619 5716 d 19 h vana /trunk/kernel/arch/ia64/src/smp/ IA64:smp.c Remove not important logs  
3617 5717 d 11 h svoboda /trunk/kernel/generic/ Fix race that allowed Udebug to report THREAD_B events for threads already reported with THREAD_READ.  
3616 5717 d 11 h vana /trunk/ IA64 correctly find CPU wakeup vector from SAL & EFI memory map support  
3614 5717 d 16 h svoboda /trunk/kernel/generic/src/syscall/ Add another stopping point.  
3613 5717 d 16 h svoboda /trunk/kernel/generic/ Merge feature from tracing: correct stopping of CPU-intensive tasks in udebug.  
3604 5718 d 5 h svoboda /trunk/kernel/generic/ Get rid of int_lock from udebug_thread_t, as is flawed.  
3603 5718 d 5 h svoboda /trunk/ Stoppability sections must be much smaller (design error). Also fix deadlock in udebug_stop(). Fix pause functionality in tracer.  
3601 5718 d 13 h svoboda /trunk/kernel/generic/ Replace 'stop' in udebug_thread_t with 'go' for consistency with nomenclature.  
3600 5718 d 14 h svoboda /trunk/kernel/generic/ Udebug comments - 'is go', not 'has go'. Missing periods.  
3594 5721 d 8 h svoboda /trunk/kernel/arch/ppc32/src/mm/ Handle page faults from userspace properly on ppc32, do not panic.  
3587 5726 d 15 h jermar /trunk/kernel/arch/ppc64/src/mm/ Remove one more occurrence of "Page fault.\n".  
3586 5726 d 15 h jermar /trunk/kernel/arch/ Do not print "Page fault.\n" if a mapping is not found during TLB refill.  
3585 5726 d 17 h svoboda /trunk/ Fix bug #67. It was obviously all my fault - I allocated r3 for pcb_ptr, which was alredy used by uarg.  
3584 5726 d 17 h svoboda /trunk/kernel/arch/ppc32/src/mm/ Bugfix: A ppc32 task would hang if it touched unmapped memory - was calling printf() with address-space mutex held.  
3583 5726 d 20 h vana /trunk/ IA64: Support for real frequency  
3581 5728 d 5 h jermar /trunk/kernel/ Some cstyle.  
3580 5728 d 6 h jermar /trunk/kernel/arch/sparc64/include/ Accesses to memory mapped registers should use volatile pointers so that
the compiler doesn't even think of trying to be clever. Add memory barriers
and fixed cstyle.
 
3578 5728 d 20 h vana /trunk/ Alfa of SMP support on IA64  
3577 5728 d 20 h vana /trunk/kernel/ Support for serial port console on IA64 as a compensation for keyboard - based on SPARC ns16550 driver  
3576 5728 d 21 h vana /trunk/kernel/generic/src/main/ On Itanium there is nead to preserve at least gp when switching APs to separate stack. Safer way is to save all context and to do it on all platforms.  
3566 5731 d 18 h svoboda /trunk/ Set meaningful names for loaded programs. Now 'tasks' kconsole command is much less obscure.  
3565 5732 d 4 h svoboda /trunk/ Fix sys_thread_create(): could fail when passed name was at end of mapping, could crash if provided name was not null-terminated.  
3522 5743 d 8 h decky /trunk/kernel/generic/include/ coding style  
3496 5748 d 7 h svoboda /trunk/kernel/generic/src/ipc/ Split and streamline kbox_thread_proc().  
3495 5748 d 8 h svoboda /trunk/kernel/generic/src/ipc/ Synchronize properly when detaching kbox from within.  
3494 5750 d 7 h svoboda /trunk/kernel/generic/src/ipc/ Fix panic when debugged task quits. Also fix memory leak when debugger disconnects first.