diff mbox series

[v4] coresight: etm4x: avoid build failure with unrolled loops

Message ID 20220614220229.1640085-1-ndesaulniers@google.com (mailing list archive)
State New, archived
Headers show
Series [v4] coresight: etm4x: avoid build failure with unrolled loops | expand

Commit Message

Nick Desaulniers June 14, 2022, 10:02 p.m. UTC
When the following configs are enabled:
* CORESIGHT
* CORESIGHT_SOURCE_ETM4X
* UBSAN
* UBSAN_TRAP

Clang fails assemble the kernel with the error:
<instantiation>:1:7: error: expected constant expression in '.inst' directive
.inst (0xd5200000|((((2) << 19) | ((1) << 16) | (((((((((((0x160 + (i * 4))))) >> 2))) >> 7) & 0x7)) << 12) | ((((((((((0x160 + (i * 4))))) >> 2))) & 0xf)) << 8) | (((((((((((0x160 + (i * 4))))) >> 2))) >> 4) & 0x7)) << 5)))|(.L__reg_num_x8))
      ^
drivers/hwtracing/coresight/coresight-etm4x-core.c:702:4: note: while in macro instantiation
                        etm4x_relaxed_read32(csa, TRCCNTVRn(i));
                        ^
drivers/hwtracing/coresight/coresight-etm4x.h:403:4: note: expanded from macro 'etm4x_relaxed_read32'
                 read_etm4x_sysreg_offset((offset), false)))
                 ^
drivers/hwtracing/coresight/coresight-etm4x.h:383:12: note: expanded from macro 'read_etm4x_sysreg_offset'
                        __val = read_etm4x_sysreg_const_offset((offset));       \
                                ^
drivers/hwtracing/coresight/coresight-etm4x.h:149:2: note: expanded from macro 'read_etm4x_sysreg_const_offset'
        READ_ETM4x_REG(ETM4x_OFFSET_TO_REG(offset))
        ^
drivers/hwtracing/coresight/coresight-etm4x.h:144:2: note: expanded from macro 'READ_ETM4x_REG'
        read_sysreg_s(ETM4x_REG_NUM_TO_SYSREG((reg)))
        ^
arch/arm64/include/asm/sysreg.h:1108:15: note: expanded from macro 'read_sysreg_s'
        asm volatile(__mrs_s("%0", r) : "=r" (__val));                  \
                     ^
arch/arm64/include/asm/sysreg.h:1074:2: note: expanded from macro '__mrs_s'
"       mrs_s " v ", " __stringify(r) "\n"                      \
 ^

Consider the definitions of TRCSSCSRn and TRCCNTVRn:
drivers/hwtracing/coresight/coresight-etm4x.h:56
 #define TRCCNTVRn(n)      (0x160 + (n * 4))
drivers/hwtracing/coresight/coresight-etm4x.h:81
 #define TRCSSCSRn(n)      (0x2A0 + (n * 4))

Where the macro parameter is expanded to i; a loop induction variable
from etm4_disable_hw.

When any compiler can determine that loops may be unrolled, then the
__builtin_constant_p check in read_etm4x_sysreg_offset() defined in
drivers/hwtracing/coresight/coresight-etm4x.h may evaluate to true. This
can lead to the expression `(0x160 + (i * 4))` being passed to
read_etm4x_sysreg_const_offset. Via the trace above, this is passed
through READ_ETM4x_REG, read_sysreg_s, and finally to __mrs_s where it
is string-ified and used directly in inline asm.

Regardless of compiler or compiler options determine whether a loop can
or can't be unrolled, which determines whether __builtin_constant_p
evaluates to true when passed an expression using a loop induction
variable, it is NEVER safe to allow the preprocessor to construction
inline asm like:
  asm volatile (".inst (0x160 + (i * 4))" : "=r"(__val));
                                 ^ expected constant expression

Replace unsafe uses of calls to etm4x_relaxed_read32 with
csdev_access_read32 when the parameter is an expression that would be
invalid inline asm so that it does not depend on the ability of the
compiler to optimize __builtin_constant_p of the expression to true.
Only when the second parameter of etm4x_relaxed_read32 expands to an
expression dependent on a loop induction variable do we need to fix
this.

This is not a bug in clang; it's a potentially unsafe use of the macro
arguments in read_etm4x_sysreg_offset dependent on __builtin_constant_p.

Link: https://github.com/ClangBuiltLinux/linux/issues/1310
Suggested-by: Arnd Bergmann <arnd@kernel.org>
Suggested-by: Tao Zhang <quic_taozha@quicinc.com>
Suggested-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Signed-off-by: Nick Desaulniers <ndesaulniers@google.com>
---
V1 (Arnd):
https://lore.kernel.org/lkml/20210225094324.3542511-1-arnd@kernel.org/

V2 (Arnd):
https://lore.kernel.org/lkml/20210429145752.3218324-1-arnd@kernel.org/

V3 (Tao):
https://lore.kernel.org/lkml/1632652550-26048-1-git-send-email-quic_taozha@quicinc.com/

 drivers/hwtracing/coresight/coresight-etm4x-core.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

Comments

Suzuki K Poulose June 23, 2022, 9:46 a.m. UTC | #1
Hi Nick

Thanks for resending the patch. Unfortunately, this one still doesn't
address my comments in the v3 [0].


On 14/06/2022 23:02, Nick Desaulniers wrote:
> When the following configs are enabled:
> * CORESIGHT
> * CORESIGHT_SOURCE_ETM4X
> * UBSAN
> * UBSAN_TRAP
> 
> Clang fails assemble the kernel with the error:
> <instantiation>:1:7: error: expected constant expression in '.inst' directive
> .inst (0xd5200000|((((2) << 19) | ((1) << 16) | (((((((((((0x160 + (i * 4))))) >> 2))) >> 7) & 0x7)) << 12) | ((((((((((0x160 + (i * 4))))) >> 2))) & 0xf)) << 8) | (((((((((((0x160 + (i * 4))))) >> 2))) >> 4) & 0x7)) << 5)))|(.L__reg_num_x8))
>        ^
> drivers/hwtracing/coresight/coresight-etm4x-core.c:702:4: note: while in macro instantiation
>                          etm4x_relaxed_read32(csa, TRCCNTVRn(i));
>                          ^
> drivers/hwtracing/coresight/coresight-etm4x.h:403:4: note: expanded from macro 'etm4x_relaxed_read32'
>                   read_etm4x_sysreg_offset((offset), false)))
>                   ^
> drivers/hwtracing/coresight/coresight-etm4x.h:383:12: note: expanded from macro 'read_etm4x_sysreg_offset'
>                          __val = read_etm4x_sysreg_const_offset((offset));       \
>                                  ^
> drivers/hwtracing/coresight/coresight-etm4x.h:149:2: note: expanded from macro 'read_etm4x_sysreg_const_offset'
>          READ_ETM4x_REG(ETM4x_OFFSET_TO_REG(offset))
>          ^
> drivers/hwtracing/coresight/coresight-etm4x.h:144:2: note: expanded from macro 'READ_ETM4x_REG'
>          read_sysreg_s(ETM4x_REG_NUM_TO_SYSREG((reg)))
>          ^
> arch/arm64/include/asm/sysreg.h:1108:15: note: expanded from macro 'read_sysreg_s'
>          asm volatile(__mrs_s("%0", r) : "=r" (__val));                  \
>                       ^
> arch/arm64/include/asm/sysreg.h:1074:2: note: expanded from macro '__mrs_s'
> "       mrs_s " v ", " __stringify(r) "\n"                      \
>   ^
> 
> Consider the definitions of TRCSSCSRn and TRCCNTVRn:
> drivers/hwtracing/coresight/coresight-etm4x.h:56
>   #define TRCCNTVRn(n)      (0x160 + (n * 4))
> drivers/hwtracing/coresight/coresight-etm4x.h:81
>   #define TRCSSCSRn(n)      (0x2A0 + (n * 4))
> 
> Where the macro parameter is expanded to i; a loop induction variable
> from etm4_disable_hw.
> 
> When any compiler can determine that loops may be unrolled, then the
> __builtin_constant_p check in read_etm4x_sysreg_offset() defined in
> drivers/hwtracing/coresight/coresight-etm4x.h may evaluate to true. This
> can lead to the expression `(0x160 + (i * 4))` being passed to
> read_etm4x_sysreg_const_offset. Via the trace above, this is passed
> through READ_ETM4x_REG, read_sysreg_s, and finally to __mrs_s where it
> is string-ified and used directly in inline asm.
> 
> Regardless of compiler or compiler options determine whether a loop can
> or can't be unrolled, which determines whether __builtin_constant_p
> evaluates to true when passed an expression using a loop induction
> variable, it is NEVER safe to allow the preprocessor to construction
> inline asm like:
>    asm volatile (".inst (0x160 + (i * 4))" : "=r"(__val));
>                                   ^ expected constant expression
> 
> Replace unsafe uses of calls to etm4x_relaxed_read32 with
> csdev_access_read32 when the parameter is an expression that would be
> invalid inline asm so that it does not depend on the ability of the
> compiler to optimize __builtin_constant_p of the expression to true.
> Only when the second parameter of etm4x_relaxed_read32 expands to an
> expression dependent on a loop induction variable do we need to fix
> this.
> 
> This is not a bug in clang; it's a potentially unsafe use of the macro
> arguments in read_etm4x_sysreg_offset dependent on __builtin_constant_p.
> 
> Link: https://github.com/ClangBuiltLinux/linux/issues/1310
> Suggested-by: Arnd Bergmann <arnd@kernel.org>
> Suggested-by: Tao Zhang <quic_taozha@quicinc.com>
> Suggested-by: Suzuki K Poulose <suzuki.poulose@arm.com>
> Signed-off-by: Nick Desaulniers <ndesaulniers@google.com>
> ---
> V1 (Arnd):
> https://lore.kernel.org/lkml/20210225094324.3542511-1-arnd@kernel.org/
> 
> V2 (Arnd):
> https://lore.kernel.org/lkml/20210429145752.3218324-1-arnd@kernel.org/
> 
> V3 (Tao):
> https://lore.kernel.org/lkml/1632652550-26048-1-git-send-email-quic_taozha@quicinc.com/


Even with this patch applied, we have many more instances left
to convert. Please could you update the patch to convert all of these ?

$ grep "(i)"  coresight-etm4x* | grep -v csdev_
coresight-etm4x-core.c: * Check if TRCSSPCICRn(i) is implemented for a 
given instance.
coresight-etm4x-core.c:         etm4x_relaxed_write32(csa, 
config->seq_ctrl[i], TRCSEQEVRn(i));
coresight-etm4x-core.c:         etm4x_relaxed_write32(csa, 
config->cntrldvr[i], TRCCNTRLDVRn(i));
coresight-etm4x-core.c:         etm4x_relaxed_write32(csa, 
config->cntr_ctrl[i], TRCCNTCTLRn(i));
coresight-etm4x-core.c:         etm4x_relaxed_write32(csa, 
config->cntr_val[i], TRCCNTVRn(i));
coresight-etm4x-core.c:         etm4x_relaxed_write32(csa, 
config->res_ctrl[i], TRCRSCTLRn(i));
coresight-etm4x-core.c:         etm4x_relaxed_write32(csa, 
config->ss_ctrl[i], TRCSSCCRn(i));
coresight-etm4x-core.c:         etm4x_relaxed_write32(csa, 
config->ss_status[i], TRCSSCSRn(i));
coresight-etm4x-core.c:                 etm4x_relaxed_write32(csa, 
config->ss_pe_cmp[i], TRCSSPCICRn(i));
coresight-etm4x-core.c:         etm4x_relaxed_write64(csa, 
config->addr_val[i], TRCACVRn(i));
coresight-etm4x-core.c:         etm4x_relaxed_write64(csa, 
config->addr_acc[i], TRCACATRn(i));
coresight-etm4x-core.c:         etm4x_relaxed_write64(csa, 
config->ctxid_pid[i], TRCCIDCVRn(i));
coresight-etm4x-core.c:         etm4x_relaxed_write64(csa, 
config->vmid_val[i], TRCVMIDCVRn(i));
coresight-etm4x-core.c:         state->trcseqevr[i] = etm4x_read32(csa, 
TRCSEQEVRn(i));
coresight-etm4x-core.c:         state->trccntrldvr[i] = 
etm4x_read32(csa, TRCCNTRLDVRn(i));
coresight-etm4x-core.c:         state->trccntctlr[i] = etm4x_read32(csa, 
TRCCNTCTLRn(i));
coresight-etm4x-core.c:         state->trccntvr[i] = etm4x_read32(csa, 
TRCCNTVRn(i));
coresight-etm4x-core.c:         state->trcrsctlr[i] = etm4x_read32(csa, 
TRCRSCTLRn(i));
coresight-etm4x-core.c:         state->trcssccr[i] = etm4x_read32(csa, 
TRCSSCCRn(i));
coresight-etm4x-core.c:         state->trcsscsr[i] = etm4x_read32(csa, 
TRCSSCSRn(i));
coresight-etm4x-core.c:                 state->trcsspcicr[i] = 
etm4x_read32(csa, TRCSSPCICRn(i));
coresight-etm4x-core.c:         state->trcacvr[i] = etm4x_read64(csa, 
TRCACVRn(i));
coresight-etm4x-core.c:         state->trcacatr[i] = etm4x_read64(csa, 
TRCACATRn(i));
coresight-etm4x-core.c:         state->trccidcvr[i] = etm4x_read64(csa, 
TRCCIDCVRn(i));
coresight-etm4x-core.c:         state->trcvmidcvr[i] = etm4x_read64(csa, 
TRCVMIDCVRn(i));
coresight-etm4x-core.c:         etm4x_relaxed_write32(csa, 
state->trcseqevr[i], TRCSEQEVRn(i));
coresight-etm4x-core.c:         etm4x_relaxed_write32(csa, 
state->trccntrldvr[i], TRCCNTRLDVRn(i));
coresight-etm4x-core.c:         etm4x_relaxed_write32(csa, 
state->trccntctlr[i], TRCCNTCTLRn(i));
coresight-etm4x-core.c:         etm4x_relaxed_write32(csa, 
state->trccntvr[i], TRCCNTVRn(i));
coresight-etm4x-core.c:         etm4x_relaxed_write32(csa, 
state->trcrsctlr[i], TRCRSCTLRn(i));
coresight-etm4x-core.c:         etm4x_relaxed_write32(csa, 
state->trcssccr[i], TRCSSCCRn(i));
coresight-etm4x-core.c:         etm4x_relaxed_write32(csa, 
state->trcsscsr[i], TRCSSCSRn(i));
coresight-etm4x-core.c:                 etm4x_relaxed_write32(csa, 
state->trcsspcicr[i], TRCSSPCICRn(i));
coresight-etm4x-core.c:         etm4x_relaxed_write64(csa, 
state->trcacvr[i], TRCACVRn(i));
coresight-etm4x-core.c:         etm4x_relaxed_write64(csa, 
state->trcacatr[i], TRCACATRn(i));
coresight-etm4x-core.c:         etm4x_relaxed_write64(csa, 
state->trccidcvr[i], TRCCIDCVRn(i));
coresight-etm4x-core.c:         etm4x_relaxed_write64(csa, 
state->trcvmidcvr[i], TRCVMIDCVRn(i));

[0] 
https://lore.kernel.org/lkml/48162555-2a67-60bc-ea4b-8720e7b98a22@arm.com/


Kind regards
Suzuki

> 
>   drivers/hwtracing/coresight/coresight-etm4x-core.c | 6 +++---
>   1 file changed, 3 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/hwtracing/coresight/coresight-etm4x-core.c b/drivers/hwtracing/coresight/coresight-etm4x-core.c
> index 87299e99dabb..7c6bd85e36d4 100644
> --- a/drivers/hwtracing/coresight/coresight-etm4x-core.c
> +++ b/drivers/hwtracing/coresight/coresight-etm4x-core.c
> @@ -836,13 +836,13 @@ static void etm4_disable_hw(void *info)
>   	/* read the status of the single shot comparators */
>   	for (i = 0; i < drvdata->nr_ss_cmp; i++) {
>   		config->ss_status[i] =
> -			etm4x_relaxed_read32(csa, TRCSSCSRn(i));
> +			csdev_access_read32(csa, TRCSSCSRn(i));
>   	}
>   
>   	/* read back the current counter values */
>   	for (i = 0; i < drvdata->nr_cntr; i++) {
>   		config->cntr_val[i] =
> -			etm4x_relaxed_read32(csa, TRCCNTVRn(i));
> +			csdev_access_read32(csa, TRCCNTVRn(i));
>   	}
>   
>   	coresight_disclaim_device_unlocked(csdev);
> @@ -1177,7 +1177,7 @@ static void etm4_init_arch_data(void *info)
>   	drvdata->nr_ss_cmp = FIELD_GET(TRCIDR4_NUMSSCC_MASK, etmidr4);
>   	for (i = 0; i < drvdata->nr_ss_cmp; i++) {
>   		drvdata->config.ss_status[i] =
> -			etm4x_relaxed_read32(csa, TRCSSCSRn(i));
> +			csdev_access_read32(csa, TRCSSCSRn(i));
>   	}
>   	/* NUMCIDC, bits[27:24] number of Context ID comparators for tracing */
>   	drvdata->numcidc = FIELD_GET(TRCIDR4_NUMCIDC_MASK, etmidr4);
diff mbox series

Patch

diff --git a/drivers/hwtracing/coresight/coresight-etm4x-core.c b/drivers/hwtracing/coresight/coresight-etm4x-core.c
index 87299e99dabb..7c6bd85e36d4 100644
--- a/drivers/hwtracing/coresight/coresight-etm4x-core.c
+++ b/drivers/hwtracing/coresight/coresight-etm4x-core.c
@@ -836,13 +836,13 @@  static void etm4_disable_hw(void *info)
 	/* read the status of the single shot comparators */
 	for (i = 0; i < drvdata->nr_ss_cmp; i++) {
 		config->ss_status[i] =
-			etm4x_relaxed_read32(csa, TRCSSCSRn(i));
+			csdev_access_read32(csa, TRCSSCSRn(i));
 	}
 
 	/* read back the current counter values */
 	for (i = 0; i < drvdata->nr_cntr; i++) {
 		config->cntr_val[i] =
-			etm4x_relaxed_read32(csa, TRCCNTVRn(i));
+			csdev_access_read32(csa, TRCCNTVRn(i));
 	}
 
 	coresight_disclaim_device_unlocked(csdev);
@@ -1177,7 +1177,7 @@  static void etm4_init_arch_data(void *info)
 	drvdata->nr_ss_cmp = FIELD_GET(TRCIDR4_NUMSSCC_MASK, etmidr4);
 	for (i = 0; i < drvdata->nr_ss_cmp; i++) {
 		drvdata->config.ss_status[i] =
-			etm4x_relaxed_read32(csa, TRCSSCSRn(i));
+			csdev_access_read32(csa, TRCSSCSRn(i));
 	}
 	/* NUMCIDC, bits[27:24] number of Context ID comparators for tracing */
 	drvdata->numcidc = FIELD_GET(TRCIDR4_NUMCIDC_MASK, etmidr4);