d/freebsd-skq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
b9e99ecd32
freebsd-skq/sys/sparc64/include/wstate.h

92 lines
4.4 KiB
C
Raw Normal View History

Implement a slightly different window spill/fill algorithm for dealing with user windows in kernel mode. We split the windows using %otherwin, but instead of spilling user window directly to the pcb, we attempt to spill to user space. If this fails because a stack page is not resident (or the stack is smashed), the fault handler at tl 2 will detect the situation and resume at tl 1 again where recovery code can spill to the pcb. Any windows that have been saved to the pcb will be copied out to the user stack on return from kernel mode. Add a first stab at 32 bit window handling. This uses much of the same recovery code as above because the alignment of the stack pointer is used to detect 32 bit code. Attempting to spill a 32 bit window to a 64 bit stack, or vice versa, will cause an alignment fault. The recovery code then changes the window state to vector to a 32 bit spill/fill handler and retries the faulting instruction. Add ktr traces in useful places during trap processing. Adjust comments to reflect new code and add many more.
2001-09-03 23:10:45 +00:00
/*-
* Copyright (c) 1997 Berkeley Software Design, Inc. All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. Berkeley Software Design Inc's name may not be used to endorse or
* promote products derived from this software without specific prior
* written permission.
*
* THIS SOFTWARE IS PROVIDED BY BERKELEY SOFTWARE DESIGN INC ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL BERKELEY SOFTWARE DESIGN INC BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*
* from: BSDI: wstate.h,v 1.4 1997/09/18 13:05:51 torek Exp
* $FreeBSD$
*/
#ifndef _MACHINE_WSTATE_H_
#define _MACHINE_WSTATE_H_
/*
* Window state register bits.
*
* There really are no bits per se, just the two fields WSTATE.NORMAL
* and WSTATE.OTHER. The rest is up to software.
*
* We use WSTATE_NORMAL to represent user mode or kernel mode saves
* (whichever is currently in effect) and WSTATE_OTHER to represent
* user mode saves (only).
*
* We use the low bit to suggest 32-bit mode, with the next bit set
* once we succeed in saving in some mode. That is, if the WSTATE_ASSUME
* bit is set, the spill or fill handler we use will be one that makes
* an assumption about the proper window-save mode. If the spill or
* fill fails with an alignment fault, the spill or fill op should
* take the `assume' bit away retry the instruction that caused the
* spill or fill. This will use the new %wstate, which will test for
* which mode to use. The alignment fault code helps us out here by
* resuming the spill vector at offset +70, where we are allowed to
* execute two instructions (i.e., write to %wstate and RETRY).
*
* If the ASSUME bit is not set when the alignment fault occurs, the
* given stack pointer is hopelessly wrong (and the spill, if it is a
* spill, should be done as a sort of "panic spill") -- so those two
* instructions will be a branch sequence.
*
* Note that locore.s assumes this same bit layout (since the translation
* from "bits" to "{spill,fill}_N_{normal,other}" is done in hardware).
*
* The value 0 is preferred for unknown to make it easy to start in
* unknown state and continue in whichever state unknown succeeds in --
* a successful "other" save, for instance, can just set %wstate to
* ASSUMExx << USERSHIFT and thus leave the kernel state "unknown".
*
* We also need values for managing the somewhat tricky transition from
* user to kernel and back, so we use the one remaining free bit to mean
* "although this looks like kernel mode, the window(s) involved are
* user windows and should be saved ASI_AIUP". Everything else is
* otherwise the same, but we need not bother with assumptions in this
* mode (we expect it to apply to at most one window spill or fill),
* i.e., WSTATE_TRANSITION can ignore WSTATE_ASSUME if it likes.
*/
Implement a nested window state. This avoids attempting to spill a user window to the user stack while in a nested kernel trap. We do this for entry to the kernel from user mode, but if we get an interrupt in kernel mode while there are still user windows in the cpu, and we attempt to spill to the user stack, we may take too many nested traps and overflow the trap stack, causing a red state exception. This is needed by upcoming changes to allow the user tsb to not be locked in the tlb. Reviewed by: tmm
2002-02-25 18:37:17 +00:00
#define WSTATE_NORMAL_MASK 1 /* wstate normal minus transition */
Rename definitions to better match the hardware wstate fields. Don't include WSTATE_TRANSITION in WSTATE_NORMAL_MASK.
2001-12-29 07:12:30 +00:00
#define WSTATE_OTHER_SHIFT 3 /* for wstate other / user */
Implement a nested window state. This avoids attempting to spill a user window to the user stack while in a nested kernel trap. We do this for entry to the kernel from user mode, but if we get an interrupt in kernel mode while there are still user windows in the cpu, and we attempt to spill to the user stack, we may take too many nested traps and overflow the trap stack, causing a red state exception. This is needed by upcoming changes to allow the user tsb to not be locked in the tlb. Reviewed by: tmm
2002-02-25 18:37:17 +00:00
#define WSTATE_OTHER_MASK /* wstate other minus nested */ \
(WSTATE_NORMAL_MASK << WSTATE_OTHER_SHIFT)
Implement a slightly different window spill/fill algorithm for dealing with user windows in kernel mode. We split the windows using %otherwin, but instead of spilling user window directly to the pcb, we attempt to spill to user space. If this fails because a stack page is not resident (or the stack is smashed), the fault handler at tl 2 will detect the situation and resume at tl 1 again where recovery code can spill to the pcb. Any windows that have been saved to the pcb will be copied out to the user stack on return from kernel mode. Add a first stab at 32 bit window handling. This uses much of the same recovery code as above because the alignment of the stack pointer is used to detect 32 bit code. Attempting to spill a 32 bit window to a 64 bit stack, or vice versa, will cause an alignment fault. The recovery code then changes the window state to vector to a 32 bit spill/fill handler and retries the faulting instruction. Add ktr traces in useful places during trap processing. Adjust comments to reflect new code and add many more.
2001-09-03 23:10:45 +00:00
Add a definition for normal kernel window state.
2001-10-20 17:08:33 +00:00
#define WSTATE_KERNEL 0 /* normal kernel wstate */
Implement a nested window state. This avoids attempting to spill a user window to the user stack while in a nested kernel trap. We do this for entry to the kernel from user mode, but if we get an interrupt in kernel mode while there are still user windows in the cpu, and we attempt to spill to the user stack, we may take too many nested traps and overflow the trap stack, causing a red state exception. This is needed by upcoming changes to allow the user tsb to not be locked in the tlb. Reviewed by: tmm
2002-02-25 18:37:17 +00:00
#define WSTATE_USER_64 0 /* normal 64bit user wstate */
#define WSTATE_USER_32 1 /* normal 32bit user wstate */
#define WSTATE_TRANSITION 2 /* if set, force user window */
#define WSTATE_NESTED /* if set, spill must not fault */ \
(WSTATE_TRANSITION << WSTATE_OTHER_SHIFT)
Add a definition for normal kernel window state.
2001-10-20 17:08:33 +00:00
Implement a slightly different window spill/fill algorithm for dealing with user windows in kernel mode. We split the windows using %otherwin, but instead of spilling user window directly to the pcb, we attempt to spill to user space. If this fails because a stack page is not resident (or the stack is smashed), the fault handler at tl 2 will detect the situation and resume at tl 1 again where recovery code can spill to the pcb. Any windows that have been saved to the pcb will be copied out to the user stack on return from kernel mode. Add a first stab at 32 bit window handling. This uses much of the same recovery code as above because the alignment of the stack pointer is used to detect 32 bit code. Attempting to spill a 32 bit window to a 64 bit stack, or vice versa, will cause an alignment fault. The recovery code then changes the window state to vector to a 32 bit spill/fill handler and retries the faulting instruction. Add ktr traces in useful places during trap processing. Adjust comments to reflect new code and add many more.
2001-09-03 23:10:45 +00:00
#endif /* !_MACHINE_WSTATE_H_ */
Reference in New Issue Copy Permalink