Series comparison

-[PATCH] kselftest/arm64: abi: fix SVCR detection
+[PATCH v2] kselftest/arm64: abi: fix SVCR detection
 ...
 instruction if FEAT_SVE and !FEAT_SME.
 Fixes: 43e3f85523e4 ("kselftest/arm64: Add SME support to syscall ABI test")
 Signed-off-by: Weizhao Ouyang <o451686892@gmail.com>
 ---
- tools/testing/selftests/arm64/abi/syscall-abi-asm.S | 2 +-
+ .../selftests/arm64/abi/syscall-abi-asm.S     | 32 +++++++++----------
-file changed, 1 insertion(+), 1 deletion(-)
+file changed, 15 insertions(+), 17 deletions(-)
 diff --git a/tools/testing/selftests/arm64/abi/syscall-abi-asm.S b/tools/testing/selftests/arm64/abi/syscall-abi-asm.S
 index XXXXXXX..XXXXXXX 100644
 --- a/tools/testing/selftests/arm64/abi/syscall-abi-asm.S
 +++ b/tools/testing/selftests/arm64/abi/syscall-abi-asm.S
 @@ -XXX,XX +XXX,XX @@ do_syscall:
      stp    x27, x28, [sp, #96]
      // Set SVCR if we're doing SME
 -    cbz    x1, 1f
++    cbz    x1, load_gpr
      adrp    x2, svcr_in
      ldr    x2, [x2, :lo12:svcr_in]
-+    cbz    x1, 1f
      msr    S3_3_C4_C2_2, x2
-:
+-1:
      // Load ZA and ZT0 if enabled - uses x12 as scratch due to SME LDR
 -    tbz    x2, #SVCR_ZA_SHIFT, 1f
 +    tbz    x2, #SVCR_ZA_SHIFT, load_gpr
      mov    w12, #0
      ldr    x2, =za_in
 -2:    _ldr_za 12, 2
 +1:    _ldr_za 12, 2
      add    x2, x2, x1
      add    x12, x12, #1
      cmp    x1, x12
 -    bne    2b
 +    bne    1b
      // ZT0
      mrs    x2, S3_0_C0_C4_5    // ID_AA64SMFR0_EL1
      ubfx    x2, x2, #ID_AA64SMFR0_EL1_SMEver_SHIFT, \
               #ID_AA64SMFR0_EL1_SMEver_WIDTH
 -    cbz    x2, 1f
 +    cbz    x2, load_gpr
      adrp    x2, zt_in
      add    x2, x2, :lo12:zt_in
      _ldr_zt 2
 -1:
 +load_gpr:
      // Load GPRs x8-x28, and save our SP/FP for later comparison
      ldr    x2, =gpr_in
      add    x2, x2, #64
@@ -XXX,XX +XXX,XX @@ do_syscall:
      str    x30, [x2], #8        // LR
      // Load FPRs if we're not doing neither SVE nor streaming SVE
 -    cbnz    x0, 1f
 +    cbnz    x0, check_sve_in
      ldr    x2, =svcr_in
 -    tbnz    x2, #SVCR_SM_SHIFT, 1f
 +    tbnz    x2, #SVCR_SM_SHIFT, check_sve_in
      ldr    x2, =fpr_in
      ldp    q0, q1, [x2]
@@ -XXX,XX +XXX,XX @@ do_syscall:
      ldp    q30, q31, [x2, #16 * 30]
      b    2f
 -1:
 +check_sve_in:
      // Load the SVE registers if we're doing SVE/SME
      ldr    x2, =z_in
@@ -XXX,XX +XXX,XX @@ do_syscall:
      stp    q30, q31, [x2, #16 * 30]
      // Save SVCR if we're doing SME
 -    cbz    x1, 1f
 +    cbz    x1, check_sve_out
      mrs    x2, S3_3_C4_C2_2
      adrp    x3, svcr_out
      str    x2, [x3, :lo12:svcr_out]
 -1:
      // Save ZA if it's enabled - uses x12 as scratch due to SME STR
 -    tbz    x2, #SVCR_ZA_SHIFT, 1f
 +    tbz    x2, #SVCR_ZA_SHIFT, check_sve_out
      mov    w12, #0
      ldr    x2, =za_out
 -2:    _str_za 12, 2
 +1:    _str_za 12, 2
      add    x2, x2, x1
      add    x12, x12, #1
      cmp    x1, x12
 -    bne    2b
 +    bne    1b
      // ZT0
      mrs    x2, S3_0_C0_C4_5    // ID_AA64SMFR0_EL1
      ubfx    x2, x2, #ID_AA64SMFR0_EL1_SMEver_SHIFT, \
              #ID_AA64SMFR0_EL1_SMEver_WIDTH
 -    cbz    x2, 1f
 +    cbz    x2, check_sve_out
      adrp    x2, zt_out
      add    x2, x2, :lo12:zt_out
      _str_zt 2
 -1:
 +check_sve_out:
      // Save the SVE state if we have some
      cbz    x0, 1f
 --
 .45.2

When using svcr_in to check ZA and Streaming Mode, we should make sure
that the value in x2 is correct, otherwise it may trigger an Illegal
instruction if FEAT_SVE and !FEAT_SME.

Fixes: 43e3f85523e4 ("kselftest/arm64: Add SME support to syscall ABI test")
Signed-off-by: Weizhao Ouyang <o451686892@gmail.com>
---
 .../selftests/arm64/abi/syscall-abi-asm.S     | 32 +++++++++----------
 1 file changed, 15 insertions(+), 17 deletions(-)

diff --git a/tools/testing/selftests/arm64/abi/syscall-abi-asm.S b/tools/testing/selftests/arm64/abi/syscall-abi-asm.S
index XXXXXXX..XXXXXXX 100644
--- a/tools/testing/selftests/arm64/abi/syscall-abi-asm.S
+++ b/tools/testing/selftests/arm64/abi/syscall-abi-asm.S
@@ -XXX,XX +XXX,XX @@ do_syscall:
 	stp	x27, x28, [sp, #96]
 
 	// Set SVCR if we're doing SME
-	cbz	x1, 1f
+	cbz	x1, load_gpr
 	adrp	x2, svcr_in
 	ldr	x2, [x2, :lo12:svcr_in]
 	msr	S3_3_C4_C2_2, x2
-1:
 
 	// Load ZA and ZT0 if enabled - uses x12 as scratch due to SME LDR
-	tbz	x2, #SVCR_ZA_SHIFT, 1f
+	tbz	x2, #SVCR_ZA_SHIFT, load_gpr
 	mov	w12, #0
 	ldr	x2, =za_in
-2:	_ldr_za 12, 2
+1:	_ldr_za 12, 2
 	add	x2, x2, x1
 	add	x12, x12, #1
 	cmp	x1, x12
-	bne	2b
+	bne	1b
 
 	// ZT0
 	mrs	x2, S3_0_C0_C4_5	// ID_AA64SMFR0_EL1
 	ubfx	x2, x2, #ID_AA64SMFR0_EL1_SMEver_SHIFT, \
 			 #ID_AA64SMFR0_EL1_SMEver_WIDTH
-	cbz	x2, 1f
+	cbz	x2, load_gpr
 	adrp	x2, zt_in
 	add	x2, x2, :lo12:zt_in
 	_ldr_zt 2
-1:
 
+load_gpr:
 	// Load GPRs x8-x28, and save our SP/FP for later comparison
 	ldr	x2, =gpr_in
 	add	x2, x2, #64
@@ -XXX,XX +XXX,XX @@ do_syscall:
 	str	x30, [x2], #8		// LR
 
 	// Load FPRs if we're not doing neither SVE nor streaming SVE
-	cbnz	x0, 1f
+	cbnz	x0, check_sve_in
 	ldr	x2, =svcr_in
-	tbnz	x2, #SVCR_SM_SHIFT, 1f
+	tbnz	x2, #SVCR_SM_SHIFT, check_sve_in
 
 	ldr	x2, =fpr_in
 	ldp	q0, q1, [x2]
@@ -XXX,XX +XXX,XX @@ do_syscall:
 	ldp	q30, q31, [x2, #16 * 30]
 
 	b	2f
-1:
 
+check_sve_in:
 	// Load the SVE registers if we're doing SVE/SME
 
 	ldr	x2, =z_in
@@ -XXX,XX +XXX,XX @@ do_syscall:
 	stp	q30, q31, [x2, #16 * 30]
 
 	// Save SVCR if we're doing SME
-	cbz	x1, 1f
+	cbz	x1, check_sve_out
 	mrs	x2, S3_3_C4_C2_2
 	adrp	x3, svcr_out
 	str	x2, [x3, :lo12:svcr_out]
-1:
 
 	// Save ZA if it's enabled - uses x12 as scratch due to SME STR
-	tbz	x2, #SVCR_ZA_SHIFT, 1f
+	tbz	x2, #SVCR_ZA_SHIFT, check_sve_out
 	mov	w12, #0
 	ldr	x2, =za_out
-2:	_str_za 12, 2
+1:	_str_za 12, 2
 	add	x2, x2, x1
 	add	x12, x12, #1
 	cmp	x1, x12
-	bne	2b
+	bne	1b
 
 	// ZT0
 	mrs	x2, S3_0_C0_C4_5	// ID_AA64SMFR0_EL1
 	ubfx	x2, x2, #ID_AA64SMFR0_EL1_SMEver_SHIFT, \
 			#ID_AA64SMFR0_EL1_SMEver_WIDTH
-	cbz	x2, 1f
+	cbz	x2, check_sve_out
 	adrp	x2, zt_out
 	add	x2, x2, :lo12:zt_out
 	_str_zt 2
-1:
 
+check_sve_out:
 	// Save the SVE state if we have some
 	cbz	x0, 1f
 
-- 
2.45.2