Re: [PATCH ipsec-next v1 6/7] bpf: selftests: test_tunnel: Disable CO-RE relocations

[Yonghong Song <yonghong.song@linux.dev>]

On 11/27/23 12:44 AM, Yonghong Song wrote:
>
> On 11/26/23 8:52 PM, Eduard Zingerman wrote:
>> On Sun, 2023-11-26 at 18:04 -0600, Daniel Xu wrote:
>> [...]
>>>> Tbh I'm not sure. This test passes with preserve_static_offset
>>>> because it suppresses preserve_access_index. In general clang
>>>> translates bitfield access to a set of IR statements like:
>>>>
>>>>    C:
>>>>      struct foo {
>>>>        unsigned _;
>>>>        unsigned a:1;
>>>>        ...
>>>>      };
>>>>      ... foo->a ...
>>>>
>>>>    IR:
>>>>      %a = getelementptr inbounds %struct.foo, ptr %0, i32 0, i32 1
>>>>      %bf.load = load i8, ptr %a, align 4
>>>>      %bf.clear = and i8 %bf.load, 1
>>>>      %bf.cast = zext i8 %bf.clear to i32
>>>>
>>>> With preserve_static_offset the getelementptr+load are replaced by a
>>>> single statement which is preserved as-is till code generation,
>>>> thus load with align 4 is preserved.
>>>>
>>>> On the other hand, I'm not sure that clang guarantees that load or
>>>> stores used for bitfield access would be always aligned according to
>>>> verifier expectations.
>>>>
>>>> I think we should check if there are some clang knobs that prevent
>>>> generation of unaligned memory access. I'll take a look.
>>> Is there a reason to prefer fixing in compiler? I'm not opposed to it,
>>> but the downside to compiler fix is it takes years to propagate and
>>> sprinkles ifdefs into the code.
>>>
>>> Would it be possible to have an analogue of BPF_CORE_READ_BITFIELD()?
>> Well, the contraption below passes verification, tunnel selftest
>> appears to work. I might have messed up some shifts in the macro, 
>> though.
>
> I didn't test it. But from high level it should work.
>
>>
>> Still, if clang would peek unlucky BYTE_{OFFSET,SIZE} for a particular
>> field access might be unaligned.
>
> clang should pick a sensible BYTE_SIZE/BYTE_OFFSET to meet
> alignment requirement. This is also required for BPF_CORE_READ_BITFIELD.
>
>>
>> ---
>>
>> diff --git a/tools/testing/selftests/bpf/progs/test_tunnel_kern.c 
>> b/tools/testing/selftests/bpf/progs/test_tunnel_kern.c
>> index 3065a716544d..41cd913ac7ff 100644
>> --- a/tools/testing/selftests/bpf/progs/test_tunnel_kern.c
>> +++ b/tools/testing/selftests/bpf/progs/test_tunnel_kern.c
>> @@ -9,6 +9,7 @@
>>   #include "vmlinux.h"
>>   #include <bpf/bpf_helpers.h>
>>   #include <bpf/bpf_endian.h>
>> +#include <bpf/bpf_core_read.h>
>>   #include "bpf_kfuncs.h"
>>   #include "bpf_tracing_net.h"
>>   @@ -144,6 +145,38 @@ int ip6gretap_get_tunnel(struct __sk_buff *skb)
>>       return TC_ACT_OK;
>>   }
>>   +#define BPF_CORE_WRITE_BITFIELD(s, field, new_val) ({            \
>> +    void *p = (void *)s + __CORE_RELO(s, field, BYTE_OFFSET);    \
>> +    unsigned byte_size = __CORE_RELO(s, field, BYTE_SIZE);        \
>> +    unsigned lshift = __CORE_RELO(s, field, LSHIFT_U64); \
>> +    unsigned rshift = __CORE_RELO(s, field, RSHIFT_U64); \
>> +    unsigned bit_size = (rshift - lshift);                \
>> +    unsigned long long nval, val, hi, lo;                \
>> +                                    \
>> +    asm volatile("" : "=r"(p) : "0"(p));                \
>
> Use asm volatile("" : "+r"(p)) ?
>
>> +                                    \
>> +    switch (byte_size) {                        \
>> +    case 1: val = *(unsigned char *)p; break;            \
>> +    case 2: val = *(unsigned short *)p; break;            \
>> +    case 4: val = *(unsigned int *)p; break;            \
>> +    case 8: val = *(unsigned long long *)p; break;            \
>> +    }                                \
>> +    hi = val >> (bit_size + rshift);                \
>> +    hi <<= bit_size + rshift;                    \
>> +    lo = val << (bit_size + lshift);                \
>> +    lo >>= bit_size + lshift;                    \
>> +    nval = new_val;                            \
>> +    nval <<= lshift;                        \
>> +    nval >>= rshift;                        \
>> +    val = hi | nval | lo;                        \
>> +    switch (byte_size) {                        \
>> +    case 1: *(unsigned char *)p      = val; break;            \
>> +    case 2: *(unsigned short *)p     = val; break;            \
>> +    case 4: *(unsigned int *)p       = val; break;            \
>> +    case 8: *(unsigned long long *)p = val; break;            \
>> +    }                                \
>> +})
>
> I think this should be put in libbpf public header files but not sure
> where to put it. bpf_core_read.h although it is core write?
>
> But on the other hand, this is a uapi struct bitfield write,
> strictly speaking, CORE write is really unnecessary here. It
> would be great if we can relieve users from dealing with
> such unnecessary CORE writes. In that sense, for this particular
> case, I would prefer rewriting the code by using byte-level
> stores...
or preserve_static_offset to clearly mean to undo bitfield CORE ...

[...]