From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mail-wr1-f47.google.com (mail-wr1-f47.google.com [209.85.221.47]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id AF7222E8B83 for ; Thu, 9 Apr 2026 14:51:19 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.221.47 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1775746282; cv=none; b=Y+v+vGqSRgJxhs5nFhKD4pQe20HGQdgzilwo1dmGg0UM2s+TP0aZbzf+uDS6GcTC3ksrLiUsoXz1PjbSd6FUNB6tlg96uvPdGINRXPcleub1SxRKnFm6azlgM3J1hWeF5doE6lmbDoGgrQHMM6qvy6YyxKWdGjaYq+yRnQFskdY= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1775746282; c=relaxed/simple; bh=egI9q0znF73vG3hvf7kbEtPWcbPku7o6x+9QloxXuns=; h=Message-ID:Date:MIME-Version:Subject:To:Cc:References:From: In-Reply-To:Content-Type; b=AJ12auDce8kfqP6Kr8/W78VjLssBvs94qAaaQdKPitEolpQQ/aZ2sX0HqjJcXgBd5C5B7htW/QbK/+7fvPuDXRSBhbqEv/Bgz424IS6Rox5l4h5zwMrKn7iGAYrx1Oz8KOEEXNPPEP+XHxL4WNWFmcMF7L/cZwPG2Yk2caNbAwo= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=gmail.com; spf=pass smtp.mailfrom=gmail.com; dkim=pass (2048-bit key) header.d=gmail.com header.i=@gmail.com header.b=HXHKEdv1; arc=none smtp.client-ip=209.85.221.47 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=gmail.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=gmail.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=gmail.com header.i=@gmail.com header.b="HXHKEdv1" Received: by mail-wr1-f47.google.com with SMTP id ffacd0b85a97d-43ccda008cdso545292f8f.0 for ; Thu, 09 Apr 2026 07:51:19 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20251104; t=1775746278; x=1776351078; darn=vger.kernel.org; h=content-transfer-encoding:in-reply-to:from:content-language :references:cc:to:subject:user-agent:mime-version:date:message-id :from:to:cc:subject:date:message-id:reply-to; bh=XCFONB0BZGVLwfaMI+ElBjgfDz4mBl9N0XIpJxn0i1E=; b=HXHKEdv1+uTx1xyiTI/LTvZFSdmop5o5vLNaRK9e796R8CAt0+xsSIWkL/ohNUMKS5 5QZm3Z7BVUA5JQcAuDlRwpUUeDdR4rVhc09pGf0EaUB5kOE193wl+pTmpUl8Tlc1XyGt rbauLCRZtpiRRf9aaZ6bPBCpsnaGMPURjRuZC0MfALhXXy7VWYjmyz65frSpZLaXtymw von3ZESHCoLMRdmRne1zLZqjCf8AEWWAAts+dBuPUTDwXJH/vbKdozeDvYvBpm7a66pk XXsQA9h5qjOVEs6/2LuwDYd4DHg8Ah/Nms1BxY2XbJxdRrk05f8PCF4DwzQOAdFN9/VL QRAg== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20251104; t=1775746278; x=1776351078; h=content-transfer-encoding:in-reply-to:from:content-language :references:cc:to:subject:user-agent:mime-version:date:message-id :x-gm-gg:x-gm-message-state:from:to:cc:subject:date:message-id :reply-to; bh=XCFONB0BZGVLwfaMI+ElBjgfDz4mBl9N0XIpJxn0i1E=; b=P7Y712Xwj0CqyBCL4zTc3ZCEZFnSYkBUO2vEk2zU4JNhUoG0fc0fZsR8K7BcbC8uYk TFTeI6ykuOxM9HBiJhtZiR/GaX3MK7tVCYC3BFGlZi2oR4tz3x68jlIkKtacUaLLQKSN To3vRzSVnh7chfVzBjl8mvbVukWXKi6R/pn9ZaYasDWP4XwwTXOKhCKAVabSe3PDSvsW yPt4J5E7cgbpxoFqkbPV7PVqLd1OWjtTQnLcsQRjW9bIZvOr9z99Wo8JUkbFJMkzmH7a kgszhfvHi5O8sBHJebJ3hu6BcjbZA+66O4WMa5A+3DzVZMf/ZlRrZ8F6BriMlLVxQ36u yFFw== X-Forwarded-Encrypted: i=1; AJvYcCWy7SyOL96URF/RM69+PHbtxjqO3CqGkxg/JtynBl27cFw5AuwhApR5rKAVycrl4ae9/MU=@vger.kernel.org X-Gm-Message-State: AOJu0YzwpHt3JMWqQ5L3PnZqsvXNoPtXJkIityLD4/2AVYYycWVy5QXu 38P3UHOaYDOuyHbQmrebPVBf8nTvnGyNGAksrhKIXVUuO4VlJ71R4HZk X-Gm-Gg: AeBDiev/ivDl6Q6an2rmFxzPOiPYY1Re3FApkPw38omq3enu5OSk844elg0+RZJA5by O3XX2sbGvbsZGk82NgHXAg14SUIv+kkIy1+mUN9q+k+EUsgbzME0HHuA+dmUUCi/YPoRc5yuKDQ cXnlygQXMlvLEgV1jYET1OeqBRurIQlkuDdmy/Tq0CgZAAUheupMeM+DrCyVFnitSLoi7fbGoZ+ aRbPsTUga9thf45tdzXVICLfYCY09K8DbFtRgc69526s/EeqauzStEx5JhRltTC02Uk+/5+IQLM Hz+NWrgr7dfjLjaAZyvq9HryDA3ZVHAclLq7D7BL2dLGG5FPKp4V/49tlyQZD7ZE6UJIEe/unbd 6gfty6ecoYUc4qn9X2Uw79uSC9urpqMJUeWko7gxFCcVKnCOl0Ki8Bm9obKxo0Dvv3ar1mPxtXO bo6JG83kFgSyYeoD5iOCPOZv2QFqzDKo2X7gmrOoN4rngVJXS4lMd655bM4eV85GJPTetiXKU/k Kx5WRHpRWU= X-Received: by 2002:a05:6000:4014:b0:43c:a352:c366 with SMTP id ffacd0b85a97d-43d595c4f98mr6133887f8f.25.1775746277373; Thu, 09 Apr 2026 07:51:17 -0700 (PDT) Received: from ?IPV6:2a02:8109:a307:d900:a778:a387:b2af:188b? ([2a02:8109:a307:d900:a778:a387:b2af:188b]) by smtp.gmail.com with ESMTPSA id ffacd0b85a97d-43d473b8c1esm28181423f8f.12.2026.04.09.07.51.16 (version=TLS1_3 cipher=TLS_AES_128_GCM_SHA256 bits=128/128); Thu, 09 Apr 2026 07:51:16 -0700 (PDT) Message-ID: Date: Thu, 9 Apr 2026 15:51:15 +0100 Precedence: bulk X-Mailing-List: bpf@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 User-Agent: Mozilla Thunderbird Subject: Re: [PATCH bpf v2 2/2] bpf: Avoid faultable build ID reads under mm locks To: Ihor Solodrai , Puranjay Mohan Cc: Andrii Nakryiko , Daniel Borkmann , Alexei Starovoitov , Song Liu , Shakeel Butt , bpf@vger.kernel.org, linux-kernel@vger.kernel.org, kernel-team@meta.com References: <20260409010604.1439087-1-ihor.solodrai@linux.dev> <20260409010604.1439087-3-ihor.solodrai@linux.dev> Content-Language: en-US From: Mykyta Yatsenko In-Reply-To: Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit On 4/9/26 5:20 AM, Ihor Solodrai wrote: > > > On 4/8/26 6:06 PM, Ihor Solodrai wrote: >> Sleepable build ID parsing can block in __kernel_read() [1], so the >> stackmap sleepable path must not call it while holding mmap_lock or a >> per-VMA read lock. >> >> The issue and the fix are conceptually similar to a recent procfs >> patch [2]. >> >> Resolve each covered VMA with a stable read-side reference, preferring >> lock_vma_under_rcu() and falling back to mmap_read_lock() only long >> enough to acquire the VMA read lock. Take a reference to the backing >> file, drop the VMA lock, and then parse the build ID through >> (sleepable) build_id_parse_file(). >> >> [1]: https://lore.kernel.org/all/20251218005818.614819-1-shakeel.butt@linux.dev/ >> [2]: https://lore.kernel.org/all/20260128183232.2854138-1-andrii@kernel.org/ >> >> Fixes: 777a8560fd29 ("lib/buildid: use __kernel_read() for sleepable context") >> Assisted-by: Codex:gpt-5.4 >> Suggested-by: Puranjay Mohan >> Signed-off-by: Ihor Solodrai >> --- >> kernel/bpf/stackmap.c | 139 ++++++++++++++++++++++++++++++++++++++++++ >> 1 file changed, 139 insertions(+) >> >> diff --git a/kernel/bpf/stackmap.c b/kernel/bpf/stackmap.c >> index 4ef0fd06cea5..de3d89e20a1e 100644 >> --- a/kernel/bpf/stackmap.c >> +++ b/kernel/bpf/stackmap.c >> @@ -9,6 +9,7 @@ >> #include >> #include >> #include >> +#include >> #include "percpu_freelist.h" >> #include "mmap_unlock_work.h" >> >> @@ -158,6 +159,139 @@ static inline void stack_map_build_id_set_ip(struct bpf_stack_build_id *id) >> memset(id->build_id, 0, BUILD_ID_SIZE_MAX); >> } >> >> +enum stack_map_vma_lock_state { >> + STACK_MAP_LOCKED_NONE = 0, >> + STACK_MAP_LOCKED_VMA, >> + STACK_MAP_LOCKED_MMAP, >> +}; >> + >> +struct stack_map_vma_lock { >> + enum stack_map_vma_lock_state state; >> + struct vm_area_struct *vma; >> + struct mm_struct *mm; >> +}; >> + >> +static struct vm_area_struct *stack_map_lock_vma(struct stack_map_vma_lock *lock, unsigned long ip) >> +{ >> + struct mm_struct *mm = lock->mm; >> + struct vm_area_struct *vma; >> + >> + if (WARN_ON_ONCE(!mm)) >> + return NULL; >> + >> + vma = lock_vma_under_rcu(mm, ip); >> + if (vma) >> + goto vma_locked; >> + >> + if (!mmap_read_trylock(mm)) >> + return NULL; >> + >> + vma = vma_lookup(mm, ip); >> + if (!vma) { >> + mmap_read_unlock(mm); >> + return NULL; >> + } >> + >> +#ifdef CONFIG_PER_VMA_LOCK > > Puranjay and others, > > I tried to test this patch with CONFIG_PER_VMA_LOCK=n, and it turns > out it's not easy to turn off. > > config PER_VMA_LOCK > def_bool y > depends on ARCH_SUPPORTS_PER_VMA_LOCK && MMU && SMP > > AFAIU this definition means that PER_VMA_LOCK is automatically true if > it's dependencies are satisfied. The only practical way to do it > appears to be disabling SMP, which is probably very unusual. > > So I am thinking, maybe instead of having to manage both mmap and vma > lock in stackmap.c, we can assume CONFIG_PER_VMA_LOCK=y, and have > > #ifndef CONFIG_PER_VMA_LOCK > return -ENOTSUPP > #endif > > somewhere in the stack_map_get_build_id_offset() callstack? > Say in __bpf_get_stackid()? > > Thoughts? > I looks like you are right, it should be safe to drop CONFIG_PER_VMA_LOCK=n case, I understand all platforms that care about BPF going to have it enabled. >> + if (!vma_start_read_locked(vma)) { >> + mmap_read_unlock(mm); >> + return NULL; >> + } >> + mmap_read_unlock(mm); >> +#else >> + lock->state = STACK_MAP_LOCKED_MMAP; >> + lock->vma = vma; >> + return vma; >> +#endif >> + >> +vma_locked: >> + lock->state = STACK_MAP_LOCKED_VMA; >> + lock->vma = vma; >> + return vma; >> +} >> + >> +static void stack_map_unlock_vma(struct stack_map_vma_lock *lock) >> +{ >> + struct vm_area_struct *vma = lock->vma; >> + struct mm_struct *mm = lock->mm; >> + >> + switch (lock->state) { >> + case STACK_MAP_LOCKED_VMA: >> + if (WARN_ON_ONCE(!vma)) >> + break; >> + vma_end_read(vma); >> + break; >> + case STACK_MAP_LOCKED_MMAP: >> + if (WARN_ON_ONCE(!mm)) >> + break; >> + mmap_read_unlock(mm); >> + break; >> + default: >> + break; >> + } >> + >> + lock->state = STACK_MAP_LOCKED_NONE; >> + lock->vma = NULL; >> +} >> + >> +static void stack_map_get_build_id_offset_sleepable(struct bpf_stack_build_id *id_offs, >> + u32 trace_nr) >> +{ >> + struct mm_struct *mm = current->mm; >> + struct stack_map_vma_lock lock = { >> + .state = STACK_MAP_LOCKED_NONE, >> + .vma = NULL, >> + .mm = mm, >> + }; >> + struct file *file, *prev_file = NULL; >> + unsigned long vm_pgoff, vm_start; >> + struct vm_area_struct *vma; >> + const char *prev_build_id; >> + u64 ip; >> + >> + for (u32 i = 0; i < trace_nr; i++) { >> + ip = READ_ONCE(id_offs[i].ip); >> + vma = stack_map_lock_vma(&lock, ip); >> + if (!vma || !vma->vm_file) { >> + stack_map_build_id_set_ip(&id_offs[i]); >> + stack_map_unlock_vma(&lock); >> + continue; >> + } >> + >> + file = vma->vm_file; >> + vm_pgoff = vma->vm_pgoff; >> + vm_start = vma->vm_start; >> + >> + if (file == prev_file) { >> + memcpy(id_offs[i].build_id, prev_build_id, BUILD_ID_SIZE_MAX); >> + stack_map_unlock_vma(&lock); >> + goto build_id_valid; >> + } >> + >> + file = get_file(file); >> + stack_map_unlock_vma(&lock); >> + >> + /* build_id_parse_file() may block on filesystem reads */ >> + if (build_id_parse_file(file, id_offs[i].build_id, NULL)) { >> + stack_map_build_id_set_ip(&id_offs[i]); >> + fput(file); >> + continue; >> + } >> + >> + if (prev_file) >> + fput(prev_file); >> + prev_file = file; >> + prev_build_id = id_offs[i].build_id; >> + >> +build_id_valid: >> + id_offs[i].offset = (vm_pgoff << PAGE_SHIFT) + ip - vm_start; >> + id_offs[i].status = BPF_STACK_BUILD_ID_VALID; >> + } >> + >> + if (prev_file) >> + fput(prev_file); >> +} >> + >> /* >> * Expects all id_offs[i].ip values to be set to correct initial IPs. >> * They will be subsequently: >> @@ -178,6 +312,11 @@ static void stack_map_get_build_id_offset(struct bpf_stack_build_id *id_offs, >> const char *prev_build_id; >> int i; >> >> + if (may_fault && has_user_ctx) { >> + stack_map_get_build_id_offset_sleepable(id_offs, trace_nr); >> + return; >> + } >> + >> /* If the irq_work is in use, fall back to report ips. Same >> * fallback is used for kernel stack (!user) on a stackmap with >> * build_id. > >