From ac506b7f2233b35f17172304255e08cabc072aad Mon Sep 17 00:00:00 2001 From: Vineet Gupta Date: Thu, 29 Oct 2015 19:36:03 +0530 Subject: ARCv2: lib: memcpy: use local symbols Otherwise perf profiles don't charge tme to memcpy Signed-off-by: Vineet Gupta --- arch/arc/lib/memcpy-archs.S | 52 ++++++++++++++++++++++----------------------- 1 file changed, 26 insertions(+), 26 deletions(-) (limited to 'arch') diff --git a/arch/arc/lib/memcpy-archs.S b/arch/arc/lib/memcpy-archs.S index 0cab0b8a57c5..f96c75edf30a 100644 --- a/arch/arc/lib/memcpy-archs.S +++ b/arch/arc/lib/memcpy-archs.S @@ -50,26 +50,26 @@ ENTRY(memcpy) ;;; if size <= 8 cmp r2, 8 - bls.d @smallchunk + bls.d @.Lsmallchunk mov.f lp_count, r2 and.f r4, r0, 0x03 rsub lp_count, r4, 4 - lpnz @aligndestination + lpnz @.Laligndestination ;; LOOP BEGIN ldb.ab r5, [r1,1] sub r2, r2, 1 stb.ab r5, [r3,1] -aligndestination: +.Laligndestination: ;;; Check the alignment of the source and.f r4, r1, 0x03 - bnz.d @sourceunaligned + bnz.d @.Lsourceunaligned ;;; CASE 0: Both source and destination are 32bit aligned ;;; Convert len to Dwords, unfold x4 lsr.f lp_count, r2, ZOLSHFT - lpnz @copy32_64bytes + lpnz @.Lcopy32_64bytes ;; LOOP START LOADX (r6, r1) PREFETCH_READ (r1) @@ -81,25 +81,25 @@ aligndestination: STOREX (r8, r3) STOREX (r10, r3) STOREX (r4, r3) -copy32_64bytes: +.Lcopy32_64bytes: and.f lp_count, r2, ZOLAND ;Last remaining 31 bytes -smallchunk: - lpnz @copyremainingbytes +.Lsmallchunk: + lpnz @.Lcopyremainingbytes ;; LOOP START ldb.ab r5, [r1,1] stb.ab r5, [r3,1] -copyremainingbytes: +.Lcopyremainingbytes: j [blink] ;;; END CASE 0 -sourceunaligned: +.Lsourceunaligned: cmp r4, 2 - beq.d @unalignedOffby2 + beq.d @.LunalignedOffby2 sub r2, r2, 1 - bhi.d @unalignedOffby3 + bhi.d @.LunalignedOffby3 ldb.ab r5, [r1, 1] ;;; CASE 1: The source is unaligned, off by 1 @@ -114,7 +114,7 @@ sourceunaligned: or r5, r5, r6 ;; Both src and dst are aligned - lpnz @copy8bytes_1 + lpnz @.Lcopy8bytes_1 ;; LOOP START ld.ab r6, [r1, 4] prefetch [r1, 28] ;Prefetch the next read location @@ -131,7 +131,7 @@ sourceunaligned: st.ab r7, [r3, 4] st.ab r9, [r3, 4] -copy8bytes_1: +.Lcopy8bytes_1: ;; Write back the remaining 16bits EXTRACT_1 (r6, r5, 16) @@ -141,14 +141,14 @@ copy8bytes_1: stb.ab r5, [r3, 1] and.f lp_count, r2, 0x07 ;Last 8bytes - lpnz @copybytewise_1 + lpnz @.Lcopybytewise_1 ;; LOOP START ldb.ab r6, [r1,1] stb.ab r6, [r3,1] -copybytewise_1: +.Lcopybytewise_1: j [blink] -unalignedOffby2: +.LunalignedOffby2: ;;; CASE 2: The source is unaligned, off by 2 ldh.ab r5, [r1, 2] sub r2, r2, 1 @@ -159,7 +159,7 @@ unalignedOffby2: #ifdef __BIG_ENDIAN__ asl.nz r5, r5, 16 #endif - lpnz @copy8bytes_2 + lpnz @.Lcopy8bytes_2 ;; LOOP START ld.ab r6, [r1, 4] prefetch [r1, 28] ;Prefetch the next read location @@ -176,7 +176,7 @@ unalignedOffby2: st.ab r7, [r3, 4] st.ab r9, [r3, 4] -copy8bytes_2: +.Lcopy8bytes_2: #ifdef __BIG_ENDIAN__ lsr.nz r5, r5, 16 @@ -184,14 +184,14 @@ copy8bytes_2: sth.ab r5, [r3, 2] and.f lp_count, r2, 0x07 ;Last 8bytes - lpnz @copybytewise_2 + lpnz @.Lcopybytewise_2 ;; LOOP START ldb.ab r6, [r1,1] stb.ab r6, [r3,1] -copybytewise_2: +.Lcopybytewise_2: j [blink] -unalignedOffby3: +.LunalignedOffby3: ;;; CASE 3: The source is unaligned, off by 3 ;;; Hence, I need to read 1byte for achieve the 32bit alignment @@ -201,7 +201,7 @@ unalignedOffby3: #ifdef __BIG_ENDIAN__ asl.ne r5, r5, 24 #endif - lpnz @copy8bytes_3 + lpnz @.Lcopy8bytes_3 ;; LOOP START ld.ab r6, [r1, 4] prefetch [r1, 28] ;Prefetch the next read location @@ -218,7 +218,7 @@ unalignedOffby3: st.ab r7, [r3, 4] st.ab r9, [r3, 4] -copy8bytes_3: +.Lcopy8bytes_3: #ifdef __BIG_ENDIAN__ lsr.nz r5, r5, 24 @@ -226,11 +226,11 @@ copy8bytes_3: stb.ab r5, [r3, 1] and.f lp_count, r2, 0x07 ;Last 8bytes - lpnz @copybytewise_3 + lpnz @.Lcopybytewise_3 ;; LOOP START ldb.ab r6, [r1,1] stb.ab r6, [r3,1] -copybytewise_3: +.Lcopybytewise_3: j [blink] END(memcpy) -- cgit v1.2.3-59-g8ed1b From 76a8c40c65a2fc9e526fcba97c0cd1a6ecbd3438 Mon Sep 17 00:00:00 2001 From: Vineet Gupta Date: Sat, 31 Oct 2015 00:47:39 +0530 Subject: ARC: remove extraneous header include Signed-off-by: Vineet Gupta --- arch/arc/plat-sim/platform.c | 1 - 1 file changed, 1 deletion(-) (limited to 'arch') diff --git a/arch/arc/plat-sim/platform.c b/arch/arc/plat-sim/platform.c index dde692812bc1..e4fe51456808 100644 --- a/arch/arc/plat-sim/platform.c +++ b/arch/arc/plat-sim/platform.c @@ -10,7 +10,6 @@ #include #include -#include /*----------------------- Machine Descriptions ------------------------------ * -- cgit v1.2.3-59-g8ed1b From 541366da6a93f52f468b408ba24ab6bb5e4fd3d8 Mon Sep 17 00:00:00 2001 From: Vineet Gupta Date: Sat, 31 Oct 2015 01:22:51 +0530 Subject: ARC: [arcompact] Handle bus error from userspace as Interrupt not exception Bus errors from userspace on ARCompact based cores are handled by core as a high priority L2 interrupt but current code treated it as interrupt Handling an interrupt like exception is certainly not going to go unnoticed. (and it worked so far as we never saw a Bus error from userspace until IPPK guys tested a DDR controller with ECC error detection etc hence needed to explicitly trigger/handle such errors) - So move mem_service exception handler from common code into ARCv2 code. - In ARCompact code, define mem_service as L2 interrupt handler which just drops down to pure kernel mode and goes of to enqueue SIGBUS Reported-by: Nelson Pereira Tested-by: Ana Martins Signed-off-by: Vineet Gupta --- arch/arc/kernel/entry-arcv2.S | 19 +++++++++++++++++++ arch/arc/kernel/entry-compact.S | 29 +++++++++++++++++++++++++---- arch/arc/kernel/entry.S | 17 ----------------- 3 files changed, 44 insertions(+), 21 deletions(-) (limited to 'arch') diff --git a/arch/arc/kernel/entry-arcv2.S b/arch/arc/kernel/entry-arcv2.S index 445e63a10754..cbfec79137bf 100644 --- a/arch/arc/kernel/entry-arcv2.S +++ b/arch/arc/kernel/entry-arcv2.S @@ -91,6 +91,25 @@ ENTRY(EV_DCError) flag 1 END(EV_DCError) +; --------------------------------------------- +; Memory Error Exception Handler +; - Unlike ARCompact, handles Bus errors for both User/Kernel mode, +; Instruction fetch or Data access, under a single Exception Vector +; --------------------------------------------- + +ENTRY(mem_service) + + EXCEPTION_PROLOGUE + + lr r0, [efa] + mov r1, sp + + FAKE_RET_FROM_EXCPN + + bl do_memory_error + b ret_from_exception +END(mem_service) + ENTRY(EV_Misaligned) EXCEPTION_PROLOGUE diff --git a/arch/arc/kernel/entry-compact.S b/arch/arc/kernel/entry-compact.S index 59f52035b4ea..431433929189 100644 --- a/arch/arc/kernel/entry-compact.S +++ b/arch/arc/kernel/entry-compact.S @@ -142,16 +142,12 @@ int1_saved_reg: .zero 4 /* Each Interrupt level needs its own scratch */ -#ifdef CONFIG_ARC_COMPACT_IRQ_LEVELS - ARCFP_DATA int2_saved_reg .type int2_saved_reg, @object .size int2_saved_reg, 4 int2_saved_reg: .zero 4 -#endif - ; --------------------------------------------- .section .text, "ax",@progbits @@ -215,6 +211,31 @@ END(handle_interrupt_level2) #endif +; --------------------------------------------- +; User Mode Memory Bus Error Interrupt Handler +; (Kernel mode memory errors handled via seperate exception vectors) +; --------------------------------------------- +ENTRY(mem_service) + + INTERRUPT_PROLOGUE 2 + + mov r0, ilink2 + mov r1, sp + + ; User process needs to be killed with SIGBUS, but first need to get + ; out of the L2 interrupt context (drop to pure kernel mode) and jump + ; off to "C" code where SIGBUS in enqueued + lr r3, [status32] + bclr r3, r3, STATUS_A2_BIT + or r3, r3, (STATUS_E1_MASK|STATUS_E2_MASK) + sr r3, [status32_l2] + mov ilink2, 1f + rtie +1: + bl do_memory_error + b ret_from_exception +END(mem_service) + ; --------------------------------------------- ; Level 1 ISR ; --------------------------------------------- diff --git a/arch/arc/kernel/entry.S b/arch/arc/kernel/entry.S index 589abf5172d6..2efb0625331d 100644 --- a/arch/arc/kernel/entry.S +++ b/arch/arc/kernel/entry.S @@ -92,23 +92,6 @@ ENTRY(instr_service) b ret_from_exception END(instr_service) -; --------------------------------------------- -; Memory Error Exception Handler -; --------------------------------------------- - -ENTRY(mem_service) - - EXCEPTION_PROLOGUE - - lr r0, [efa] - mov r1, sp - - FAKE_RET_FROM_EXCPN - - bl do_memory_error - b ret_from_exception -END(mem_service) - ; --------------------------------------------- ; Machine Check Exception Handler ; --------------------------------------------- -- cgit v1.2.3-59-g8ed1b From a6416f57ce57fb390b6ee30b12c01c29032a26af Mon Sep 17 00:00:00 2001 From: Vineet Gupta Date: Thu, 5 Nov 2015 09:13:31 +0530 Subject: ARC: use ASL assembler mnemonic ARCompact and ARCv2 only have ASL, while binutils used to support LSL as a alias mnemonic. Newer binutils (upstream) don't want to do that so replace it. Signed-off-by: Vineet Gupta --- arch/arc/mm/tlbex.S | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'arch') diff --git a/arch/arc/mm/tlbex.S b/arch/arc/mm/tlbex.S index 63860adc4814..f1967eeb32e7 100644 --- a/arch/arc/mm/tlbex.S +++ b/arch/arc/mm/tlbex.S @@ -88,7 +88,7 @@ ex_saved_reg1: #ifdef CONFIG_SMP sr r0, [ARC_REG_SCRATCH_DATA0] ; freeup r0 to code with GET_CPU_ID r0 ; get to per cpu scratch mem, - lsl r0, r0, L1_CACHE_SHIFT ; cache line wide per cpu + asl r0, r0, L1_CACHE_SHIFT ; cache line wide per cpu add r0, @ex_saved_reg1, r0 #else st r0, [@ex_saved_reg1] @@ -107,7 +107,7 @@ ex_saved_reg1: .macro TLBMISS_RESTORE_REGS #ifdef CONFIG_SMP GET_CPU_ID r0 ; get to per cpu scratch mem - lsl r0, r0, L1_CACHE_SHIFT ; each is cache line wide + asl r0, r0, L1_CACHE_SHIFT ; each is cache line wide add r0, @ex_saved_reg1, r0 ld_s r3, [r0,12] ld_s r2, [r0, 8] @@ -256,7 +256,7 @@ ex_saved_reg1: .macro CONV_PTE_TO_TLB and r3, r0, PTE_BITS_RWX ; r w x - lsl r2, r3, 3 ; Kr Kw Kx 0 0 0 (GLOBAL, kernel only) + asl r2, r3, 3 ; Kr Kw Kx 0 0 0 (GLOBAL, kernel only) and.f 0, r0, _PAGE_GLOBAL or.z r2, r2, r3 ; Kr Kw Kx Ur Uw Ux (!GLOBAL, user page) -- cgit v1.2.3-59-g8ed1b From 1cfc05cbe2c9757c6cc36804df96e3ed28e940a1 Mon Sep 17 00:00:00 2001 From: Vineet Gupta Date: Mon, 9 Nov 2015 17:48:34 +0530 Subject: ARC: cpu_relax() to be compiler barrier even for UP cpu_relax() on ARC has been barrier only for SMP (and no-op for UP). Per recent discussions, it is safer to make it a compiler barrier unconditionally. Link: http://lkml.kernel.org/r/53A7D3AA.9020100@synopsys.com Acked-by: Peter Zijlstra Signed-off-by: Vineet Gupta --- arch/arc/include/asm/processor.h | 4 ---- 1 file changed, 4 deletions(-) (limited to 'arch') diff --git a/arch/arc/include/asm/processor.h b/arch/arc/include/asm/processor.h index 44545354e9e8..1d694c1ef6d6 100644 --- a/arch/arc/include/asm/processor.h +++ b/arch/arc/include/asm/processor.h @@ -57,11 +57,7 @@ struct task_struct; * A lot of busy-wait loops in SMP are based off of non-volatile data otherwise * get optimised away by gcc */ -#ifdef CONFIG_SMP #define cpu_relax() __asm__ __volatile__ ("" : : : "memory") -#else -#define cpu_relax() do { } while (0) -#endif #define cpu_relax_lowlatency() cpu_relax() -- cgit v1.2.3-59-g8ed1b