linuxsampler-devel

[Linuxsampler-devel] problem to build linuxsampler 2.4.0 for armv7hl

[Philippe D. <phi...@la...>]

Hi

We use to create rpms of gig linuxsampler gigedit and qsampler for i686 
, x86_64 , aarch64  and  armv7hl...

There's something wrong when we try to build linuxsampler 2.4.0 for 
armv7hl :

This concerns RTMath.cpp ... We get a build failure with this message :

RTMath.cpp:79:19: error: invalid operand for instruction
     79 |     asm volatile ("mrs %0, cntvct_el0" : "=r"(t));


This is linked to the instructions from line 77 to 80 #elif 
defined(__ARM_ARCH_7A__) || defined(__ARM_ARCH_7S__) uint32_t t; asm 
volatile ("mrs %0, cntvct_el0" : "=r"(t)); return t;

Maybe we should use for ARM_ARCH_75
the same instruction as for ARMv6 ... that is :
     uint32_t t;
     asm volatile ("mrc p15, 0, %0, c15, c12, 1" : "=r" (t));
     return t;

Thanks for your work
and for your help

Philippe

Re: [Linuxsampler-devel] problem to build linuxsampler 2.4.0 for armv7hl

[Christian S. <sch...@li...>]

On Tuesday, June 17, 2025 2:16:19 AM CEST Philippe DIDIER wrote:
> Hi

Hi Philippe,

> We use to create rpms of gig linuxsampler gigedit and qsampler for i686
> , x86_64 , aarch64  and  armv7hl...
> 
> There's something wrong when we try to build linuxsampler 2.4.0 for
> armv7hl :
> 
> This concerns RTMath.cpp ... We get a build failure with this message :
> 
> RTMath.cpp:79:19: error: invalid operand for instruction
>      79 |     asm volatile ("mrs %0, cntvct_el0" : "=r"(t));
> 
> 
> This is linked to the instructions from line 77 to 80 #elif
> defined(__ARM_ARCH_7A__) || defined(__ARM_ARCH_7S__) uint32_t t; asm
> volatile ("mrs %0, cntvct_el0" : "=r"(t)); return t;
> 
> Maybe we should use for ARM_ARCH_75

Yeah, looks like that's an ARMv8 (a.k.a. ARM64) feature.

> the same instruction as for ARMv6 ... that is :
>      uint32_t t;
>      asm volatile ("mrc p15, 0, %0, c15, c12, 1" : "=r" (t));
>      return t;

I think that also needs some correction. Can you test the following patch?

Index: src/common/RTMath.cpp
===================================================================
--- src/common/RTMath.cpp       (revision 4334)
+++ src/common/RTMath.cpp       (working copy)
@@ -74,16 +74,12 @@
     uint64_t t;
     asm volatile ("mrs %0, cntvct_el0" : "=r"(t));
     return (time_stamp_t) t;
-    #elif defined(__ARM_ARCH_7A__) || defined(__ARM_ARCH_7S__)
-    uint32_t t;
-    asm volatile ("mrs %0, cntvct_el0" : "=r"(t));
-    return t;
     #elif defined(__APPLE__)
     return (time_stamp_t) mach_absolute_time();
     #elif defined(__arm__) /* ARMv6 and older */
     # warning ARM 'mrc' instruction requires special runtime privileges!
     uint32_t t;
-    asm volatile ("mrc p15, 0, %0, c15, c12, 1" : "=r" (t));
+    asm volatile ("mrc p15, 0, %0, c9, c13, 0" : "=r" (t));
     return t;
     #else // we don't want to use a slow generic solution
     #  error "Sorry, LinuxSampler lacks time stamp code for your system."

Would be good to know not only if it compiles, but whether it works at 
runtime. Because as you can see from the warning, special privileges or 
preceding system configuration steps might be required for this to work at 
all.

If not, it's getting pretty ugly for the user. Since the sampler would first 
start without complaint and eventually after loading a sound the app would die 
with illegal instruction exception.

/Christian

Re: [Linuxsampler-devel] problem to build linuxsampler 2.4.0 for armv7hl

[Philippe D. <phi...@la...>]

Le 18/06/2025 à 12:25, Christian Schoenebeck a écrit :
> On Tuesday, June 17, 2025 2:16:19 AM CEST Philippe DIDIER wrote:
>> Hi
> Hi Philippe,
>
>> We use to create rpms of gig linuxsampler gigedit and qsampler for i686
>> , x86_64 , aarch64  and  armv7hl...
>>
>> There's something wrong when we try to build linuxsampler 2.4.0 for
>> armv7hl :
>>
>> This concerns RTMath.cpp ... We get a build failure with this message :
>>
>> RTMath.cpp:79:19: error: invalid operand for instruction
>>       79 |     asm volatile ("mrs %0, cntvct_el0" : "=r"(t));
>>
>>
>> This is linked to the instructions from line 77 to 80 #elif
>> defined(__ARM_ARCH_7A__) || defined(__ARM_ARCH_7S__) uint32_t t; asm
>> volatile ("mrs %0, cntvct_el0" : "=r"(t)); return t;
>>
>> Maybe we should use for ARM_ARCH_75
> Yeah, looks like that's an ARMv8 (a.k.a. ARM64) feature.
>
>> the same instruction as for ARMv6 ... that is :
>>       uint32_t t;
>>       asm volatile ("mrc p15, 0, %0, c15, c12, 1" : "=r" (t));
>>       return t;
> I think that also needs some correction. Can you test the following patch?
>
> Index: src/common/RTMath.cpp
> ===================================================================
> --- src/common/RTMath.cpp       (revision 4334)
> +++ src/common/RTMath.cpp       (working copy)
> @@ -74,16 +74,12 @@
>       uint64_t t;
>       asm volatile ("mrs %0, cntvct_el0" : "=r"(t));
>       return (time_stamp_t) t;
> -    #elif defined(__ARM_ARCH_7A__) || defined(__ARM_ARCH_7S__)
> -    uint32_t t;
> -    asm volatile ("mrs %0, cntvct_el0" : "=r"(t));
> -    return t;
>       #elif defined(__APPLE__)
>       return (time_stamp_t) mach_absolute_time();
>       #elif defined(__arm__) /* ARMv6 and older */
>       # warning ARM 'mrc' instruction requires special runtime privileges!
>       uint32_t t;
> -    asm volatile ("mrc p15, 0, %0, c15, c12, 1" : "=r" (t));
> +    asm volatile ("mrc p15, 0, %0, c9, c13, 0" : "=r" (t));
>       return t;
>       #else // we don't want to use a slow generic solution
>       #  error "Sorry, LinuxSampler lacks time stamp code for your system."
>
> Would be good to know not only if it compiles, but whether it works at
> runtime. Because as you can see from the warning, special privileges or
> preceding system configuration steps might be required for this to work at
> all.
>
> If not, it's getting pretty ugly for the user. Since the sampler would first
> start without complaint and eventually after loading a sound the app would die
> with illegal instruction exception.
>
> /Christian
>
>
>
>
> _______________________________________________
> Linuxsampler-devel mailing list
> Lin...@li...
> https://lists.sourceforge.net/lists/listinfo/linuxsampler-devel


Hi Christian

We have tested 2 patches (the first from me the second from an other 
packager) :


--- src/common/RTMath.cpp
+++ src/common/RTMath.cpp
@@ -75,9 +75,9 @@
      asm volatile ("mrs %0, cntvct_el0" : "=r"(t));
      return (time_stamp_t) t;
      #elif defined(__ARM_ARCH_7A__) || defined(__ARM_ARCH_7S__)
-    uint32_t t;
-    asm volatile ("mrs %0, cntvct_el0" : "=r"(t));
-    return t;
+    uint32_t t;
+    asm volatile ("mrc p15, 0, %0, c15, c12, 1" : "=r" (t));
+    return t;
      #elif defined(__APPLE__)
      return (time_stamp_t) mach_absolute_time();
      #elif defined(__arm__) /* ARMv6 and older */



--- src/common/RTMath.cpp
+++ src/common/RTMath.cpp
@@ -75,9 +75,9 @@
      asm volatile ("mrs %0, cntvct_el0" : "=r"(t));
      return (time_stamp_t) t;
      #elif defined(__ARM_ARCH_7A__) || defined(__ARM_ARCH_7S__)
-    uint32_t t;
-    asm volatile ("mrs %0, cntvct_el0" : "=r"(t));
-    return t;
+    uint32_t hi, lo;
+    asm volatile ("mrrc p15, 1, %0, %1, c14" : "=r"(lo), "=r" (hi));
+    return lo;
      #elif defined(__APPLE__)
      return (time_stamp_t) mach_absolute_time();
      #elif defined(__arm__) /* ARMv6 and older */


Both of them allow to compile for armv7hl on our BuildSystem
Nevertheless I choosed to apply the first one with
return t   and not      return lo
for coherency with the other arches


I'm gonna test to compile with

asm volatile ("mrc p15, 0, %0, c9, c13, 0" : "=r" (t));


Unfortunately we can't test the use on a real armv7hl system (using a 
virtual system is not enough)