Re: yet another approach Was: [PATCH bpf-next v3 4/5] bpf, x86: Add jit support for private stack

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

On 10/3/24 10:35 AM, Alexei Starovoitov wrote:
> On Thu, Oct 3, 2024 at 6:40 AM Kumar Kartikeya Dwivedi <memxor@gmail.com> wrote:
>> On Thu, 3 Oct 2024 at 08:17, Yonghong Song <yonghong.song@linux.dev> wrote:
>>>
>>> On 10/1/24 6:26 PM, Alexei Starovoitov wrote:
>>>> On Tue, Oct 1, 2024 at 5:23 PM Kumar Kartikeya Dwivedi <memxor@gmail.com> wrote:
>>>>> Makes sense, though will we have cases where hierarchical scheduling
>>>>> attaches the same prog at different points of the hierarchy?
>>>> I'm not sure anyone was asking for such a use case.
>>>>
>>>>> Then the
>>>>> limit of 4 may not be enough (e.g. say with cgroup nested levels > 4).
>>>> Well, 4 was the number from TJ.
>>>>
>>>> Anyway the proposed pseudo code:
>>>>
>>>> __bpf_prog_enter_recur_limited()
>>>> {
>>>>     cnt = this_cpu_inc_return(*(prog->active));
>>>>     if (cnt > 4) {
>>>>        inc_miss
>>>>        return 0;
>>>>     }
>>>>    // pass cnt into bpf prog somehow, like %rdx ?
>>>>    // or re-read prog->active from prog
>>>> }
>>>>
>>>>
>>>> then in the prologue emit:
>>>>
>>>> push rbp
>>>> mov rbp, rsp
>>>> if %rdx == 1
>>>>      // main prog is called for the first time
>>>>      mov rsp, pcpu_priv_stack_top
>>>> else
>>>>      // 2+nd time main prog is called or 1+ time subprog
>>>>     sub rsp, stack_size
>>>>     if rsp < pcpu_priv_stack_bottom
>>>>       goto exit  // stack is too small, exit
>>>> fi
>>> I have tried to implement this approach (not handling
>>> recursion yet) based on the above approach. It works
>>> okay with nested bpf subprogs like
>>>      main prog  // set rsp = pcpu_priv_stack_top
>>>        subprog1 // some stack
>>>          subprog2 // some stack
>>>
>>> The pcpu_priv_stack is allocated like
>>>     priv_stack_ptr = __alloc_percpu_gfp(1024 * 16, 8, GFP_KERNEL);
>>>
>>> But whenever the prog called an external function,
>>> e.g. a helper in this case, I will get a double fault.
>>> An example could be
>>>      main prog  // set rsp = pcpu_priv_stack_top
>>>        subprog1 // some stack
>>>          subprog2 // some stack
>>>        call bpf_seq_printf
>>> (I modified bpf_iter_ipv6_route.c bpf prog for the above
>>> purpose.)
>>> I added some printk statements from the beginning of bpf_seq_printf and
>>> nothing printed out either and of course traps still happens.
>>>
>>> I tried another example without subprog and the mainprog calls
>>> a helper and the same double traps happens below too.
>>>
>>> The error log looks like
>>>
>>> [   54.024955] traps: PANIC: double fault, error_code: 0x0
>>> [   54.024969] Oops: double fault: 0000 [#1] PREEMPT SMP KASAN PTI
>>> [   54.024977] CPU: 3 UID: 0 PID: 1946 Comm: test_progs Tainted: G           OE      6.11.0-10577-gf25c172fd840-dirty #968
>>> [   54.024982] Tainted: [O]=OOT_MODULE, [E]=UNSIGNED_MODULE
>>> [   54.024983] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.14.0-0-g155821a1990b-prebuilt.qemu.org 04/01/2014
>>> [   54.024986] RIP: 0010:error_entry+0x1e/0x140
>>> [   54.024996] Code: ff ff 90 90 90 90 90 90 90 90 90 90 56 48 8b 74 24 08 48 89 7c 24 08 52 51 50 41 50 41 51 41 52 41 53 53 55 41 54 41 55 41 56 <41> 57 56 31 f6 31 d1
>>> [   54.024999] RSP: 0018:ffffe8ffff580000 EFLAGS: 00010806
>>> [   54.025002] RAX: f3f3f300f1f1f1f1 RBX: fffff91fffeb0044 RCX: ffffffff84201701
>>> [   54.025005] RDX: fffff91fffeb0044 RSI: ffffffff8420128d RDI: ffffe8ffff580178
>>> [   54.025007] RBP: ffffe8ffff580140 R08: 0000000000000000 R09: 0000000000000000
>>> [   54.025009] R10: 0000000000000000 R11: 0000000000000000 R12: dffffc0000000000
>>> [   54.025010] R13: 1ffffd1fffeb0014 R14: 0000000000000003 R15: ffffe8ffff580178
>>> [   54.025012] FS:  00007fd076525d00(0000) GS:ffff8881f7180000(0000) knlGS:0000000000000000
>>> [   54.025015] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
>>> [   54.025017] CR2: ffffe8ffff57fff8 CR3: 000000010cd80002 CR4: 0000000000370ef0
>>> [   54.025021] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
>>> [   54.025022] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
>>> [   54.025024] Call Trace:
>>> [   54.025026]  <#DF>
>>> [   54.025028]  ? __die_body+0xaf/0xc0
>>> [   54.025032]  ? die+0x2f/0x50
>>> [   54.025036]  ? exc_double_fault+0x73/0x80
>>> [   54.025040]  ? asm_exc_double_fault+0x23/0x30
>>> [   54.025044]  ? common_interrupt_return+0xb1/0xcc
>>> [   54.025048]  ? asm_exc_page_fault+0xd/0x30
>>> [   54.025051]  ? error_entry+0x1e/0x140
>>> [   54.025055]  </#DF>
>>> [   54.025056] Modules linked in: bpf_testmod(OE)
>>> [   54.025061] ---[ end trace 0000000000000000 ]---
>>>
>>> Maybe somebody could give a hint why I got a double fault
>>> when calling external functions (outside of bpf programs)
>>> with allocated stack?
>>>
>> I will help in debugging. Can you share the patch you applied locally
>> so I can reproduce?
> Looks like the idea needs more thought.
>
> in_task_stack() won't recognize the private stack,
> so it will look like stack overflow and double fault.
>
> do you have CONFIG_VMAP_STACK ?

Yes, my above test runs fine withCONFIG_VMAP_STACK. Let me guard private stack support with 
CONFIG_VMAP_STACK for now. Not sure whether distributions enable 
CONFIG_VMAP_STACK or not.