mirrors/linux-stable
1
0
Fork 0
mirror of https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git synced 2024-10-11 19:19:42 +00:00
Code Issues Releases Wiki Activity
linux-stable/arch/powerpc/kernel/vector.S

329 lines
6.6 KiB
ArmAsm
Raw Normal View History

powerpc: Move VMX and VSX asm code to vector.S Currently, load_up_altivec and give_up_altivec are duplicated in 32-bit and 64-bit. This creates a common implementation that is moved away from head_32.S, head_64.S and misc_64.S and into vector.S, using the same macros we already use for our common implementation of load_up_fpu. I also moved the VSX code over to vector.S though in that case I didn't make it build on 32-bit (yet). Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-06-02 21:17:37 +00:00
#include <asm/processor.h>
powerpc: Merge enough to start building in arch/powerpc. This creates the directory structure under arch/powerpc and a bunch of Kconfig files. It does a first-cut merge of arch/powerpc/mm, arch/powerpc/lib and arch/powerpc/platforms/powermac. This is enough to build a 32-bit powermac kernel with ARCH=powerpc. For now we are getting some unmerged files from arch/ppc/kernel and arch/ppc/syslib, or arch/ppc64/kernel. This makes some minor changes to files in those directories and files outside arch/powerpc. The boot directory is still not merged. That's going to be interesting. Signed-off-by: Paul Mackerras <paulus@samba.org>
2005-09-26 06:04:21 +00:00
#include <asm/ppc_asm.h>
powerpc: Use reg.h instead of processor.h when we just want reg names Now that the register names and bit definitions are all in reg.h, use that instead of processor.h in assembly code in a few places. Signed-off-by: Paul Mackerras <paulus@samba.org>
2005-10-10 12:20:10 +00:00
#include <asm/reg.h>
powerpc: Move VMX and VSX asm code to vector.S Currently, load_up_altivec and give_up_altivec are duplicated in 32-bit and 64-bit. This creates a common implementation that is moved away from head_32.S, head_64.S and misc_64.S and into vector.S, using the same macros we already use for our common implementation of load_up_fpu. I also moved the VSX code over to vector.S though in that case I didn't make it build on 32-bit (yet). Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-06-02 21:17:37 +00:00
#include <asm/asm-offsets.h>
#include <asm/cputable.h>
#include <asm/thread_info.h>
#include <asm/page.h>
powerpc: Remove second definition of STACK_FRAME_OVERHEAD Since STACK_FRAME_OVERHEAD is defined in asm/ptrace.h and that is ASSEMBER safe, we can just include that instead of going via asm-offsets.h. Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2010-11-18 15:06:17 +00:00
#include <asm/ptrace.h>
ppc: move exports to definitions Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2016-01-14 04:33:46 +00:00
#include <asm/export.h>
powerpc: Move VMX and VSX asm code to vector.S Currently, load_up_altivec and give_up_altivec are duplicated in 32-bit and 64-bit. This creates a common implementation that is moved away from head_32.S, head_64.S and misc_64.S and into vector.S, using the same macros we already use for our common implementation of load_up_fpu. I also moved the VSX code over to vector.S though in that case I didn't make it build on 32-bit (yet). Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-06-02 21:17:37 +00:00
powerpc: Provide for giveup_fpu/altivec to save state in alternate location This provides a facility which is intended for use by KVM, where the contents of the FP/VSX and VMX (Altivec) registers can be saved away to somewhere other than the thread_struct when kernel code wants to use floating point or VMX instructions. This is done by providing a pointer in the thread_struct to indicate where the state should be saved to. The giveup_fpu() and giveup_altivec() functions test these pointers and save state to the indicated location if they are non-NULL. Note that the MSR_FP/VEC bits in task->thread.regs->msr are still used to indicate whether the CPU register state is live, even when an alternate save location is being used. This also provides load_fp_state() and load_vr_state() functions, which load up FP/VSX and VMX state from memory into the CPU registers, and corresponding store_fp_state() and store_vr_state() functions, which store FP/VSX and VMX state into memory from the CPU registers. Signed-off-by: Paul Mackerras <paulus@samba.org> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2013-09-10 10:21:10 +00:00
/*
* Load state from memory into VMX registers including VSCR.
* Assumes the caller has enabled VMX in the MSR.
*/
_GLOBAL(load_vr_state)
li r4,VRSTATE_VSCR
powerpc: Change vrX register defines to vX to match gcc and glibc As our various loops (copy, string, crypto etc) get more complicated, we want to share implementations between userspace (eg glibc) and the kernel. We also want to write userspace test harnesses to put in tools/testing/selftest. One gratuitous difference between userspace and the kernel is the VMX register definitions - the kernel uses vrX whereas both gcc and glibc use vX. Change the kernel to match userspace. Signed-off-by: Anton Blanchard <anton@samba.org> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2015-02-09 22:51:22 +00:00
lvx v0,r4,r3
mtvscr v0
powerpc: Provide for giveup_fpu/altivec to save state in alternate location This provides a facility which is intended for use by KVM, where the contents of the FP/VSX and VMX (Altivec) registers can be saved away to somewhere other than the thread_struct when kernel code wants to use floating point or VMX instructions. This is done by providing a pointer in the thread_struct to indicate where the state should be saved to. The giveup_fpu() and giveup_altivec() functions test these pointers and save state to the indicated location if they are non-NULL. Note that the MSR_FP/VEC bits in task->thread.regs->msr are still used to indicate whether the CPU register state is live, even when an alternate save location is being used. This also provides load_fp_state() and load_vr_state() functions, which load up FP/VSX and VMX state from memory into the CPU registers, and corresponding store_fp_state() and store_vr_state() functions, which store FP/VSX and VMX state into memory from the CPU registers. Signed-off-by: Paul Mackerras <paulus@samba.org> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2013-09-10 10:21:10 +00:00
REST_32VRS(0,r4,r3)
blr
ppc: move exports to definitions Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2016-01-14 04:33:46 +00:00
EXPORT_SYMBOL(load_vr_state)
powerpc: Provide for giveup_fpu/altivec to save state in alternate location This provides a facility which is intended for use by KVM, where the contents of the FP/VSX and VMX (Altivec) registers can be saved away to somewhere other than the thread_struct when kernel code wants to use floating point or VMX instructions. This is done by providing a pointer in the thread_struct to indicate where the state should be saved to. The giveup_fpu() and giveup_altivec() functions test these pointers and save state to the indicated location if they are non-NULL. Note that the MSR_FP/VEC bits in task->thread.regs->msr are still used to indicate whether the CPU register state is live, even when an alternate save location is being used. This also provides load_fp_state() and load_vr_state() functions, which load up FP/VSX and VMX state from memory into the CPU registers, and corresponding store_fp_state() and store_vr_state() functions, which store FP/VSX and VMX state into memory from the CPU registers. Signed-off-by: Paul Mackerras <paulus@samba.org> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2013-09-10 10:21:10 +00:00
/*
* Store VMX state into memory, including VSCR.
* Assumes the caller has enabled VMX in the MSR.
*/
_GLOBAL(store_vr_state)
SAVE_32VRS(0, r4, r3)
powerpc: Change vrX register defines to vX to match gcc and glibc As our various loops (copy, string, crypto etc) get more complicated, we want to share implementations between userspace (eg glibc) and the kernel. We also want to write userspace test harnesses to put in tools/testing/selftest. One gratuitous difference between userspace and the kernel is the VMX register definitions - the kernel uses vrX whereas both gcc and glibc use vX. Change the kernel to match userspace. Signed-off-by: Anton Blanchard <anton@samba.org> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2015-02-09 22:51:22 +00:00
mfvscr v0
powerpc: Provide for giveup_fpu/altivec to save state in alternate location This provides a facility which is intended for use by KVM, where the contents of the FP/VSX and VMX (Altivec) registers can be saved away to somewhere other than the thread_struct when kernel code wants to use floating point or VMX instructions. This is done by providing a pointer in the thread_struct to indicate where the state should be saved to. The giveup_fpu() and giveup_altivec() functions test these pointers and save state to the indicated location if they are non-NULL. Note that the MSR_FP/VEC bits in task->thread.regs->msr are still used to indicate whether the CPU register state is live, even when an alternate save location is being used. This also provides load_fp_state() and load_vr_state() functions, which load up FP/VSX and VMX state from memory into the CPU registers, and corresponding store_fp_state() and store_vr_state() functions, which store FP/VSX and VMX state into memory from the CPU registers. Signed-off-by: Paul Mackerras <paulus@samba.org> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2013-09-10 10:21:10 +00:00
li r4, VRSTATE_VSCR
powerpc: Change vrX register defines to vX to match gcc and glibc As our various loops (copy, string, crypto etc) get more complicated, we want to share implementations between userspace (eg glibc) and the kernel. We also want to write userspace test harnesses to put in tools/testing/selftest. One gratuitous difference between userspace and the kernel is the VMX register definitions - the kernel uses vrX whereas both gcc and glibc use vX. Change the kernel to match userspace. Signed-off-by: Anton Blanchard <anton@samba.org> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2015-02-09 22:51:22 +00:00
stvx v0, r4, r3
powerpc: Provide for giveup_fpu/altivec to save state in alternate location This provides a facility which is intended for use by KVM, where the contents of the FP/VSX and VMX (Altivec) registers can be saved away to somewhere other than the thread_struct when kernel code wants to use floating point or VMX instructions. This is done by providing a pointer in the thread_struct to indicate where the state should be saved to. The giveup_fpu() and giveup_altivec() functions test these pointers and save state to the indicated location if they are non-NULL. Note that the MSR_FP/VEC bits in task->thread.regs->msr are still used to indicate whether the CPU register state is live, even when an alternate save location is being used. This also provides load_fp_state() and load_vr_state() functions, which load up FP/VSX and VMX state from memory into the CPU registers, and corresponding store_fp_state() and store_vr_state() functions, which store FP/VSX and VMX state into memory from the CPU registers. Signed-off-by: Paul Mackerras <paulus@samba.org> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2013-09-10 10:21:10 +00:00
blr
ppc: move exports to definitions Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2016-01-14 04:33:46 +00:00
EXPORT_SYMBOL(store_vr_state)
powerpc: Provide for giveup_fpu/altivec to save state in alternate location This provides a facility which is intended for use by KVM, where the contents of the FP/VSX and VMX (Altivec) registers can be saved away to somewhere other than the thread_struct when kernel code wants to use floating point or VMX instructions. This is done by providing a pointer in the thread_struct to indicate where the state should be saved to. The giveup_fpu() and giveup_altivec() functions test these pointers and save state to the indicated location if they are non-NULL. Note that the MSR_FP/VEC bits in task->thread.regs->msr are still used to indicate whether the CPU register state is live, even when an alternate save location is being used. This also provides load_fp_state() and load_vr_state() functions, which load up FP/VSX and VMX state from memory into the CPU registers, and corresponding store_fp_state() and store_vr_state() functions, which store FP/VSX and VMX state into memory from the CPU registers. Signed-off-by: Paul Mackerras <paulus@samba.org> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2013-09-10 10:21:10 +00:00
powerpc: Move VMX and VSX asm code to vector.S Currently, load_up_altivec and give_up_altivec are duplicated in 32-bit and 64-bit. This creates a common implementation that is moved away from head_32.S, head_64.S and misc_64.S and into vector.S, using the same macros we already use for our common implementation of load_up_fpu. I also moved the VSX code over to vector.S though in that case I didn't make it build on 32-bit (yet). Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-06-02 21:17:37 +00:00
/*
* Disable VMX for the task which had it previously,
* and save its vector registers in its thread_struct.
* Enables the VMX for use in the kernel on return.
* On SMP we know the VMX is free, since we give it up every
* switch (ie, no lazy save of the vector registers).
powerpc: Don't corrupt user registers on 32-bit Commit de79f7b9f6 ("powerpc: Put FP/VSX and VR state into structures") modified load_up_fpu() and load_up_altivec() in such a way that they now use r7 and r8. Unfortunately, the callers of these functions on 32-bit machines then return to userspace via fast_exception_return, which doesn't restore all of the volatile GPRs, but only r1, r3 -- r6 and r9 -- r12. This was causing userspace segfaults and other userspace misbehaviour on 32-bit machines. This fixes the problem by changing the register usage of load_up_fpu() and load_up_altivec() to avoid using r7 and r8 and instead use r6 and r10. This also adds comments to those functions saying which registers may be used. Signed-off-by: Paul Mackerras <paulus@samba.org> Tested-by: Scott Wood <scottwood@freescale.com> (on e500mc, so no altivec) Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2013-10-23 08:40:02 +00:00
*
* Note that on 32-bit this can only use registers that will be
* restored by fast_exception_return, i.e. r3 - r6, r10 and r11.
powerpc: Move VMX and VSX asm code to vector.S Currently, load_up_altivec and give_up_altivec are duplicated in 32-bit and 64-bit. This creates a common implementation that is moved away from head_32.S, head_64.S and misc_64.S and into vector.S, using the same macros we already use for our common implementation of load_up_fpu. I also moved the VSX code over to vector.S though in that case I didn't make it build on 32-bit (yet). Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-06-02 21:17:37 +00:00
*/
_GLOBAL(load_up_altivec)
mfmsr r5 /* grab the current MSR */
oris r5,r5,MSR_VEC@h
MTMSRD(r5) /* enable use of AltiVec now */
isync
powerpc: Improve comment explaining why we modify VRSAVE The comment explaining why we modify VRSAVE is misleading, glibc does rely on the behaviour. Update the comment. Signed-off-by: Anton Blanchard <anton@samba.org> Reviewed-by: Cyril Bur <cyrilbur@gmail.com> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2016-05-19 18:41:34 +00:00
/*
* While userspace in general ignores VRSAVE, glibc uses it as a boolean
* to optimise userspace context save/restore. Whenever we take an
* altivec unavailable exception we must set VRSAVE to something non
* zero. Set it to all 1s. See also the programming note in the ISA.
powerpc: Move VMX and VSX asm code to vector.S Currently, load_up_altivec and give_up_altivec are duplicated in 32-bit and 64-bit. This creates a common implementation that is moved away from head_32.S, head_64.S and misc_64.S and into vector.S, using the same macros we already use for our common implementation of load_up_fpu. I also moved the VSX code over to vector.S though in that case I didn't make it build on 32-bit (yet). Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-06-02 21:17:37 +00:00
*/
mfspr r4,SPRN_VRSAVE
powerpc: Fix usage of 64-bit instruction in 32-bit altivec code e821ea70f3b4873b50056a1e0f74befed1014c09 introduced a bug by copying some 64-bit originated code as-is to be used by both 32 and 64-bit but this code contains a 64-bit ony "cmpdi" instruction. This changes it to cmpwi, which is fine since VRSAVE can only contains a 32-bit value anyway. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> CC: <stable@kernel.org>
2009-12-08 18:45:45 +00:00
cmpwi 0,r4,0
powerpc: Move VMX and VSX asm code to vector.S Currently, load_up_altivec and give_up_altivec are duplicated in 32-bit and 64-bit. This creates a common implementation that is moved away from head_32.S, head_64.S and misc_64.S and into vector.S, using the same macros we already use for our common implementation of load_up_fpu. I also moved the VSX code over to vector.S though in that case I didn't make it build on 32-bit (yet). Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-06-02 21:17:37 +00:00
bne+ 1f
li r4,-1
mtspr SPRN_VRSAVE,r4
1:
/* enable use of VMX after return */
#ifdef CONFIG_PPC32
powerpc: Use names rather than numbers for SPRGs (v2) The kernel uses SPRG registers for various purposes, typically in low level assembly code as scratch registers or to hold per-cpu global infos such as the PACA or the current thread_info pointer. We want to be able to easily shuffle the usage of those registers as some implementations have specific constraints realted to some of them, for example, some have userspace readable aliases, etc.. and the current choice isn't always the best. This patch should not change any code generation, and replaces the usage of SPRN_SPRGn everywhere in the kernel with a named replacement and adds documentation next to the definition of the names as to what those are used for on each processor family. The only parts that still use the original numbers are bits of KVM or suspend/resume code that just blindly needs to save/restore all the SPRGs. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-07-14 20:52:54 +00:00
mfspr r5,SPRN_SPRG_THREAD /* current task's THREAD (phys) */
powerpc: Move VMX and VSX asm code to vector.S Currently, load_up_altivec and give_up_altivec are duplicated in 32-bit and 64-bit. This creates a common implementation that is moved away from head_32.S, head_64.S and misc_64.S and into vector.S, using the same macros we already use for our common implementation of load_up_fpu. I also moved the VSX code over to vector.S though in that case I didn't make it build on 32-bit (yet). Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-06-02 21:17:37 +00:00
oris r9,r9,MSR_VEC@h
#else
ld r4,PACACURRENT(r13)
addi r5,r4,THREAD /* Get THREAD */
oris r12,r12,MSR_VEC@h
std r12,_MSR(r1)
#endif
powerpc: Restore FPU/VEC/VSX if previously used Currently the FPU, VEC and VSX facilities are lazily loaded. This is not a problem unless a process is using these facilities. Modern versions of GCC are very good at automatically vectorising code, new and modernised workloads make use of floating point and vector facilities, even the kernel makes use of vectorised memcpy. All this combined greatly increases the cost of a syscall since the kernel uses the facilities sometimes even in syscall fast-path making it increasingly common for a thread to take an *_unavailable exception soon after a syscall, not to mention potentially taking all three. The obvious overcompensation to this problem is to simply always load all the facilities on every exit to userspace. Loading up all FPU, VEC and VSX registers every time can be expensive and if a workload does avoid using them, it should not be forced to incur this penalty. An 8bit counter is used to detect if the registers have been used in the past and the registers are always loaded until the value wraps to back to zero. Several versions of the assembly in entry_64.S were tested: 1. Always calling C. 2. Performing a common case check and then calling C. 3. A complex check in asm. After some benchmarking it was determined that avoiding C in the common case is a performance benefit (option 2). The full check in asm (option 3) greatly complicated that codepath for a negligible performance gain and the trade-off was deemed not worth it. Signed-off-by: Cyril Bur <cyrilbur@gmail.com> [mpe: Move load_vec in the struct to fill an existing hole, reword change log] Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> fixup
2016-02-29 06:53:47 +00:00
/* Don't care if r4 overflows, this is desired behaviour */
lbz r4,THREAD_LOAD_VEC(r5)
addi r4,r4,1
stb r4,THREAD_LOAD_VEC(r5)
powerpc: Don't corrupt user registers on 32-bit Commit de79f7b9f6 ("powerpc: Put FP/VSX and VR state into structures") modified load_up_fpu() and load_up_altivec() in such a way that they now use r7 and r8. Unfortunately, the callers of these functions on 32-bit machines then return to userspace via fast_exception_return, which doesn't restore all of the volatile GPRs, but only r1, r3 -- r6 and r9 -- r12. This was causing userspace segfaults and other userspace misbehaviour on 32-bit machines. This fixes the problem by changing the register usage of load_up_fpu() and load_up_altivec() to avoid using r7 and r8 and instead use r6 and r10. This also adds comments to those functions saying which registers may be used. Signed-off-by: Paul Mackerras <paulus@samba.org> Tested-by: Scott Wood <scottwood@freescale.com> (on e500mc, so no altivec) Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2013-10-23 08:40:02 +00:00
addi r6,r5,THREAD_VRSTATE
powerpc: Move VMX and VSX asm code to vector.S Currently, load_up_altivec and give_up_altivec are duplicated in 32-bit and 64-bit. This creates a common implementation that is moved away from head_32.S, head_64.S and misc_64.S and into vector.S, using the same macros we already use for our common implementation of load_up_fpu. I also moved the VSX code over to vector.S though in that case I didn't make it build on 32-bit (yet). Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-06-02 21:17:37 +00:00
li r4,1
powerpc: Put FP/VSX and VR state into structures This creates new 'thread_fp_state' and 'thread_vr_state' structures to store FP/VSX state (including FPSCR) and Altivec/VSX state (including VSCR), and uses them in the thread_struct. In the thread_fp_state, the FPRs and VSRs are represented as u64 rather than double, since we rarely perform floating-point computations on the values, and this will enable the structures to be used in KVM code as well. Similarly FPSCR is now a u64 rather than a structure of two 32-bit values. This takes the offsets out of the macros such as SAVE_32FPRS, REST_32FPRS, etc. This enables the same macros to be used for normal and transactional state, enabling us to delete the transactional versions of the macros. This also removes the unused do_load_up_fpu and do_load_up_altivec, which were in fact buggy since they didn't create large enough stack frames to account for the fact that load_up_fpu and load_up_altivec are not designed to be called from C and assume that their caller's stack frame is an interrupt frame. Signed-off-by: Paul Mackerras <paulus@samba.org> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2013-09-10 10:20:42 +00:00
li r10,VRSTATE_VSCR
powerpc: Move VMX and VSX asm code to vector.S Currently, load_up_altivec and give_up_altivec are duplicated in 32-bit and 64-bit. This creates a common implementation that is moved away from head_32.S, head_64.S and misc_64.S and into vector.S, using the same macros we already use for our common implementation of load_up_fpu. I also moved the VSX code over to vector.S though in that case I didn't make it build on 32-bit (yet). Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-06-02 21:17:37 +00:00
stw r4,THREAD_USED_VR(r5)
powerpc: Change vrX register defines to vX to match gcc and glibc As our various loops (copy, string, crypto etc) get more complicated, we want to share implementations between userspace (eg glibc) and the kernel. We also want to write userspace test harnesses to put in tools/testing/selftest. One gratuitous difference between userspace and the kernel is the VMX register definitions - the kernel uses vrX whereas both gcc and glibc use vX. Change the kernel to match userspace. Signed-off-by: Anton Blanchard <anton@samba.org> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2015-02-09 22:51:22 +00:00
lvx v0,r10,r6
mtvscr v0
powerpc: Don't corrupt user registers on 32-bit Commit de79f7b9f6 ("powerpc: Put FP/VSX and VR state into structures") modified load_up_fpu() and load_up_altivec() in such a way that they now use r7 and r8. Unfortunately, the callers of these functions on 32-bit machines then return to userspace via fast_exception_return, which doesn't restore all of the volatile GPRs, but only r1, r3 -- r6 and r9 -- r12. This was causing userspace segfaults and other userspace misbehaviour on 32-bit machines. This fixes the problem by changing the register usage of load_up_fpu() and load_up_altivec() to avoid using r7 and r8 and instead use r6 and r10. This also adds comments to those functions saying which registers may be used. Signed-off-by: Paul Mackerras <paulus@samba.org> Tested-by: Scott Wood <scottwood@freescale.com> (on e500mc, so no altivec) Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2013-10-23 08:40:02 +00:00
REST_32VRS(0,r4,r6)
powerpc: Move VMX and VSX asm code to vector.S Currently, load_up_altivec and give_up_altivec are duplicated in 32-bit and 64-bit. This creates a common implementation that is moved away from head_32.S, head_64.S and misc_64.S and into vector.S, using the same macros we already use for our common implementation of load_up_fpu. I also moved the VSX code over to vector.S though in that case I didn't make it build on 32-bit (yet). Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-06-02 21:17:37 +00:00
/* restore registers and return */
blr
/*
powerpc: Add the ability to save Altivec without giving it up This patch adds the ability to be able to save the VEC registers to the thread struct without giving up (disabling the facility) next time the process returns to userspace. This patch builds on a previous optimisation for the FPU registers in the thread copy path to avoid a possibly pointless reload of VEC state. Signed-off-by: Cyril Bur <cyrilbur@gmail.com> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2016-02-29 06:53:50 +00:00
* save_altivec(tsk)
* Save the vector registers to its thread_struct
powerpc: Move VMX and VSX asm code to vector.S Currently, load_up_altivec and give_up_altivec are duplicated in 32-bit and 64-bit. This creates a common implementation that is moved away from head_32.S, head_64.S and misc_64.S and into vector.S, using the same macros we already use for our common implementation of load_up_fpu. I also moved the VSX code over to vector.S though in that case I didn't make it build on 32-bit (yet). Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-06-02 21:17:37 +00:00
*/
powerpc: Add the ability to save Altivec without giving it up This patch adds the ability to be able to save the VEC registers to the thread struct without giving up (disabling the facility) next time the process returns to userspace. This patch builds on a previous optimisation for the FPU registers in the thread copy path to avoid a possibly pointless reload of VEC state. Signed-off-by: Cyril Bur <cyrilbur@gmail.com> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2016-02-29 06:53:50 +00:00
_GLOBAL(save_altivec)
powerpc: Move VMX and VSX asm code to vector.S Currently, load_up_altivec and give_up_altivec are duplicated in 32-bit and 64-bit. This creates a common implementation that is moved away from head_32.S, head_64.S and misc_64.S and into vector.S, using the same macros we already use for our common implementation of load_up_fpu. I also moved the VSX code over to vector.S though in that case I didn't make it build on 32-bit (yet). Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-06-02 21:17:37 +00:00
addi r3,r3,THREAD /* want THREAD of task */
powerpc: Provide for giveup_fpu/altivec to save state in alternate location This provides a facility which is intended for use by KVM, where the contents of the FP/VSX and VMX (Altivec) registers can be saved away to somewhere other than the thread_struct when kernel code wants to use floating point or VMX instructions. This is done by providing a pointer in the thread_struct to indicate where the state should be saved to. The giveup_fpu() and giveup_altivec() functions test these pointers and save state to the indicated location if they are non-NULL. Note that the MSR_FP/VEC bits in task->thread.regs->msr are still used to indicate whether the CPU register state is live, even when an alternate save location is being used. This also provides load_fp_state() and load_vr_state() functions, which load up FP/VSX and VMX state from memory into the CPU registers, and corresponding store_fp_state() and store_vr_state() functions, which store FP/VSX and VMX state into memory from the CPU registers. Signed-off-by: Paul Mackerras <paulus@samba.org> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2013-09-10 10:21:10 +00:00
PPC_LL r7,THREAD_VRSAVEAREA(r3)
powerpc: Move VMX and VSX asm code to vector.S Currently, load_up_altivec and give_up_altivec are duplicated in 32-bit and 64-bit. This creates a common implementation that is moved away from head_32.S, head_64.S and misc_64.S and into vector.S, using the same macros we already use for our common implementation of load_up_fpu. I also moved the VSX code over to vector.S though in that case I didn't make it build on 32-bit (yet). Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-06-02 21:17:37 +00:00
PPC_LL r5,PT_REGS(r3)
powerpc: Provide for giveup_fpu/altivec to save state in alternate location This provides a facility which is intended for use by KVM, where the contents of the FP/VSX and VMX (Altivec) registers can be saved away to somewhere other than the thread_struct when kernel code wants to use floating point or VMX instructions. This is done by providing a pointer in the thread_struct to indicate where the state should be saved to. The giveup_fpu() and giveup_altivec() functions test these pointers and save state to the indicated location if they are non-NULL. Note that the MSR_FP/VEC bits in task->thread.regs->msr are still used to indicate whether the CPU register state is live, even when an alternate save location is being used. This also provides load_fp_state() and load_vr_state() functions, which load up FP/VSX and VMX state from memory into the CPU registers, and corresponding store_fp_state() and store_vr_state() functions, which store FP/VSX and VMX state into memory from the CPU registers. Signed-off-by: Paul Mackerras <paulus@samba.org> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2013-09-10 10:21:10 +00:00
PPC_LCMPI 0,r7,0
bne 2f
addi r7,r3,THREAD_VRSTATE
powerpc: Add the ability to save Altivec without giving it up This patch adds the ability to be able to save the VEC registers to the thread struct without giving up (disabling the facility) next time the process returns to userspace. This patch builds on a previous optimisation for the FPU registers in the thread copy path to avoid a possibly pointless reload of VEC state. Signed-off-by: Cyril Bur <cyrilbur@gmail.com> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2016-02-29 06:53:50 +00:00
2: SAVE_32VRS(0,r4,r7)
powerpc: Change vrX register defines to vX to match gcc and glibc As our various loops (copy, string, crypto etc) get more complicated, we want to share implementations between userspace (eg glibc) and the kernel. We also want to write userspace test harnesses to put in tools/testing/selftest. One gratuitous difference between userspace and the kernel is the VMX register definitions - the kernel uses vrX whereas both gcc and glibc use vX. Change the kernel to match userspace. Signed-off-by: Anton Blanchard <anton@samba.org> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2015-02-09 22:51:22 +00:00
mfvscr v0
powerpc: Put FP/VSX and VR state into structures This creates new 'thread_fp_state' and 'thread_vr_state' structures to store FP/VSX state (including FPSCR) and Altivec/VSX state (including VSCR), and uses them in the thread_struct. In the thread_fp_state, the FPRs and VSRs are represented as u64 rather than double, since we rarely perform floating-point computations on the values, and this will enable the structures to be used in KVM code as well. Similarly FPSCR is now a u64 rather than a structure of two 32-bit values. This takes the offsets out of the macros such as SAVE_32FPRS, REST_32FPRS, etc. This enables the same macros to be used for normal and transactional state, enabling us to delete the transactional versions of the macros. This also removes the unused do_load_up_fpu and do_load_up_altivec, which were in fact buggy since they didn't create large enough stack frames to account for the fact that load_up_fpu and load_up_altivec are not designed to be called from C and assume that their caller's stack frame is an interrupt frame. Signed-off-by: Paul Mackerras <paulus@samba.org> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2013-09-10 10:20:42 +00:00
li r4,VRSTATE_VSCR
powerpc: Change vrX register defines to vX to match gcc and glibc As our various loops (copy, string, crypto etc) get more complicated, we want to share implementations between userspace (eg glibc) and the kernel. We also want to write userspace test harnesses to put in tools/testing/selftest. One gratuitous difference between userspace and the kernel is the VMX register definitions - the kernel uses vrX whereas both gcc and glibc use vX. Change the kernel to match userspace. Signed-off-by: Anton Blanchard <anton@samba.org> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2015-02-09 22:51:22 +00:00
stvx v0,r4,r7
powerpc: Move VMX and VSX asm code to vector.S Currently, load_up_altivec and give_up_altivec are duplicated in 32-bit and 64-bit. This creates a common implementation that is moved away from head_32.S, head_64.S and misc_64.S and into vector.S, using the same macros we already use for our common implementation of load_up_fpu. I also moved the VSX code over to vector.S though in that case I didn't make it build on 32-bit (yet). Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-06-02 21:17:37 +00:00
blr
#ifdef CONFIG_VSX
#ifdef CONFIG_PPC32
#error This asm code isn't ready for 32-bit kernels
#endif
/*
* load_up_vsx(unused, unused, tsk)
* Disable VSX for the task which had it previously,
* and save its vector registers in its thread_struct.
* Reuse the fp and vsx saves, but first check to see if they have
* been saved already.
*/
_GLOBAL(load_up_vsx)
/* Load FP and VSX registers if they haven't been done yet */
andi. r5,r12,MSR_FP
beql+ load_up_fpu /* skip if already loaded */
andis. r5,r12,MSR_VEC@h
beql+ load_up_altivec /* skip if already loaded */
ld r4,PACACURRENT(r13)
addi r4,r4,THREAD /* Get THREAD */
li r6,1
stw r6,THREAD_USED_VSR(r4) /* ... also set thread used vsr */
/* enable use of VSX after return */
oris r12,r12,MSR_VSX@h
std r12,_MSR(r1)
b fast_exception_return
#endif /* CONFIG_VSX */
powerpc: Merge enough to start building in arch/powerpc. This creates the directory structure under arch/powerpc and a bunch of Kconfig files. It does a first-cut merge of arch/powerpc/mm, arch/powerpc/lib and arch/powerpc/platforms/powermac. This is enough to build a 32-bit powermac kernel with ARCH=powerpc. For now we are getting some unmerged files from arch/ppc/kernel and arch/ppc/syslib, or arch/ppc64/kernel. This makes some minor changes to files in those directories and files outside arch/powerpc. The boot directory is still not merged. That's going to be interesting. Signed-off-by: Paul Mackerras <paulus@samba.org>
2005-09-26 06:04:21 +00:00
/*
* The routines below are in assembler so we can closely control the
* usage of floating-point registers. These routines must be called
* with preempt disabled.
*/
#ifdef CONFIG_PPC32
.data
fpzero:
.long 0
fpone:
.long 0x3f800000 /* 1.0 in single-precision FP */
fphalf:
.long 0x3f000000 /* 0.5 in single-precision FP */
#define LDCONST(fr, name) \
lis r11,name@ha; \
lfs fr,name@l(r11)
#else
.section ".toc","aw"
fpzero:
.tc FD_0_0[TC],0
fpone:
.tc FD_3ff00000_0[TC],0x3ff0000000000000 /* 1.0 */
fphalf:
.tc FD_3fe00000_0[TC],0x3fe0000000000000 /* 0.5 */
#define LDCONST(fr, name) \
lfd fr,name@toc(r2)
#endif
.text
/*
* Internal routine to enable floating point and set FPSCR to 0.
* Don't call it from C; it doesn't use the normal calling convention.
*/
fpenable:
#ifdef CONFIG_PPC32
stwu r1,-64(r1)
#else
stdu r1,-64(r1)
#endif
mfmsr r10
ori r11,r10,MSR_FP
mtmsr r11
isync
stfd fr0,24(r1)
stfd fr1,16(r1)
stfd fr31,8(r1)
LDCONST(fr1, fpzero)
mffs fr31
[POWERPC] 64bit FPSCR support Forthcoming machines will extend the FPSCR to 64 bits. We already had a 64-bit save area for the FPSCR, but we need to use a new form of the mtfsf instruction. Fortunately this new form is decoded as an ordinary mtfsf by existing 64-bit processors. Signed-off-by: Anton Blanchard <anton@samba.org> Signed-off-by: Paul Mackerras <paulus@samba.org>
2006-06-10 10:18:39 +00:00
MTFSF_L(fr1)
powerpc: Merge enough to start building in arch/powerpc. This creates the directory structure under arch/powerpc and a bunch of Kconfig files. It does a first-cut merge of arch/powerpc/mm, arch/powerpc/lib and arch/powerpc/platforms/powermac. This is enough to build a 32-bit powermac kernel with ARCH=powerpc. For now we are getting some unmerged files from arch/ppc/kernel and arch/ppc/syslib, or arch/ppc64/kernel. This makes some minor changes to files in those directories and files outside arch/powerpc. The boot directory is still not merged. That's going to be interesting. Signed-off-by: Paul Mackerras <paulus@samba.org>
2005-09-26 06:04:21 +00:00
blr
fpdisable:
mtlr r12
[POWERPC] 64bit FPSCR support Forthcoming machines will extend the FPSCR to 64 bits. We already had a 64-bit save area for the FPSCR, but we need to use a new form of the mtfsf instruction. Fortunately this new form is decoded as an ordinary mtfsf by existing 64-bit processors. Signed-off-by: Anton Blanchard <anton@samba.org> Signed-off-by: Paul Mackerras <paulus@samba.org>
2006-06-10 10:18:39 +00:00
MTFSF_L(fr31)
powerpc: Merge enough to start building in arch/powerpc. This creates the directory structure under arch/powerpc and a bunch of Kconfig files. It does a first-cut merge of arch/powerpc/mm, arch/powerpc/lib and arch/powerpc/platforms/powermac. This is enough to build a 32-bit powermac kernel with ARCH=powerpc. For now we are getting some unmerged files from arch/ppc/kernel and arch/ppc/syslib, or arch/ppc64/kernel. This makes some minor changes to files in those directories and files outside arch/powerpc. The boot directory is still not merged. That's going to be interesting. Signed-off-by: Paul Mackerras <paulus@samba.org>
2005-09-26 06:04:21 +00:00
lfd fr31,8(r1)
lfd fr1,16(r1)
lfd fr0,24(r1)
mtmsr r10
isync
addi r1,r1,64
blr
/*
* Vector add, floating point.
*/
_GLOBAL(vaddfp)
mflr r12
bl fpenable
li r0,4
mtctr r0
li r6,0
1: lfsx fr0,r4,r6
lfsx fr1,r5,r6
fadds fr0,fr0,fr1
stfsx fr0,r3,r6
addi r6,r6,4
bdnz 1b
b fpdisable
/*
* Vector subtract, floating point.
*/
_GLOBAL(vsubfp)
mflr r12
bl fpenable
li r0,4
mtctr r0
li r6,0
1: lfsx fr0,r4,r6
lfsx fr1,r5,r6
fsubs fr0,fr0,fr1
stfsx fr0,r3,r6
addi r6,r6,4
bdnz 1b
b fpdisable
/*
* Vector multiply and add, floating point.
*/
_GLOBAL(vmaddfp)
mflr r12
bl fpenable
stfd fr2,32(r1)
li r0,4
mtctr r0
li r7,0
1: lfsx fr0,r4,r7
lfsx fr1,r5,r7
lfsx fr2,r6,r7
fmadds fr0,fr0,fr2,fr1
stfsx fr0,r3,r7
addi r7,r7,4
bdnz 1b
lfd fr2,32(r1)
b fpdisable
/*
* Vector negative multiply and subtract, floating point.
*/
_GLOBAL(vnmsubfp)
mflr r12
bl fpenable
stfd fr2,32(r1)
li r0,4
mtctr r0
li r7,0
1: lfsx fr0,r4,r7
lfsx fr1,r5,r7
lfsx fr2,r6,r7
fnmsubs fr0,fr0,fr2,fr1
stfsx fr0,r3,r7
addi r7,r7,4
bdnz 1b
lfd fr2,32(r1)
b fpdisable
/*
* Vector reciprocal estimate. We just compute 1.0/x.
* r3 -> destination, r4 -> source.
*/
_GLOBAL(vrefp)
mflr r12
bl fpenable
li r0,4
LDCONST(fr1, fpone)
mtctr r0
li r6,0
1: lfsx fr0,r4,r6
fdivs fr0,fr1,fr0
stfsx fr0,r3,r6
addi r6,r6,4
bdnz 1b
b fpdisable
/*
* Vector reciprocal square-root estimate, floating point.
* We use the frsqrte instruction for the initial estimate followed
* by 2 iterations of Newton-Raphson to get sufficient accuracy.
* r3 -> destination, r4 -> source.
*/
_GLOBAL(vrsqrtefp)
mflr r12
bl fpenable
stfd fr2,32(r1)
stfd fr3,40(r1)
stfd fr4,48(r1)
stfd fr5,56(r1)
li r0,4
LDCONST(fr4, fpone)
LDCONST(fr5, fphalf)
mtctr r0
li r6,0
1: lfsx fr0,r4,r6
frsqrte fr1,fr0 /* r = frsqrte(s) */
fmuls fr3,fr1,fr0 /* r * s */
fmuls fr2,fr1,fr5 /* r * 0.5 */
fnmsubs fr3,fr1,fr3,fr4 /* 1 - s * r * r */
fmadds fr1,fr2,fr3,fr1 /* r = r + 0.5 * r * (1 - s * r * r) */
fmuls fr3,fr1,fr0 /* r * s */
fmuls fr2,fr1,fr5 /* r * 0.5 */
fnmsubs fr3,fr1,fr3,fr4 /* 1 - s * r * r */
fmadds fr1,fr2,fr3,fr1 /* r = r + 0.5 * r * (1 - s * r * r) */
stfsx fr1,r3,r6
addi r6,r6,4
bdnz 1b
lfd fr5,56(r1)
lfd fr4,48(r1)
lfd fr3,40(r1)
lfd fr2,32(r1)
b fpdisable
Reference in a new issue Copy permalink