{ "schema_version": "1.6.1", "id": "RPI-CVE-2024-26656", "modified": "2026-10-09T11:39:44Z", "upstream": [ "CVE-2024-26656" ], "published": "2024-04-02T06:08:57Z", "summary": "drm/amdgpu: fix use-after-free bug", "details": "In the Linux kernel, the following vulnerability has been resolved:\n\ndrm/amdgpu: fix use-after-free bug\n\nThe bug can be triggered by sending a single amdgpu_gem_userptr_ioctl\nto the AMDGPU DRM driver on any ASICs with an invalid address and size.\nThe bug was reported by Joonkyo Jung .\nFor example the following code:\n\nstatic void Syzkaller1(int fd)\n{\n\tstruct drm_amdgpu_gem_userptr arg;\n\tint ret;\n\n\targ.addr = 0xffffffffffff0000;\n\targ.size = 0x80000000; /*2 Gb*/\n\targ.flags = 0x7;\n\tret = drmIoctl(fd, 0xc1186451/*amdgpu_gem_userptr_ioctl*/, &arg);\n}\n\nDue to the address and size are not valid there is a failure in\namdgpu_hmm_register->mmu_interval_notifier_insert->__mmu_interval_notifier_insert->\ncheck_shl_overflow, but we even the amdgpu_hmm_register failure we still call\namdgpu_hmm_unregister into amdgpu_gem_object_free which causes access to a bad address.\nThe following stack is below when the issue is reproduced when Kazan is enabled:\n\n[ +0.000014] Hardware name: ASUS System Product Name/ROG STRIX B550-F GAMING (WI-FI), BIOS 1401 12/03/2020\n[ +0.000009] RIP: 0010:mmu_interval_notifier_remove+0x327/0x340\n[ +0.000017] Code: ff ff 49 89 44 24 08 48 b8 00 01 00 00 00 00 ad de 4c 89 f7 49 89 47 40 48 83 c0 22 49 89 47 48 e8 ce d1 2d 01 e9 32 ff ff ff <0f> 0b e9 16 ff ff ff 4c 89 ef e8 fa 14 b3 ff e9 36 ff ff ff e8 80\n[ +0.000014] RSP: 0018:ffffc90002657988 EFLAGS: 00010246\n[ +0.000013] RAX: 0000000000000000 RBX: 1ffff920004caf35 RCX: ffffffff8160565b\n[ +0.000011] RDX: dffffc0000000000 RSI: 0000000000000004 RDI: ffff8881a9f78260\n[ +0.000010] RBP: ffffc90002657a70 R08: 0000000000000001 R09: fffff520004caf25\n[ +0.000010] R10: 0000000000000003 R11: ffffffff8161d1d6 R12: ffff88810e988c00\n[ +0.000010] R13: ffff888126fb5a00 R14: ffff88810e988c0c R15: ffff8881a9f78260\n[ +0.000011] FS: 00007ff9ec848540(0000) GS:ffff8883cc880000(0000) knlGS:0000000000000000\n[ +0.000012] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033\n[ +0.000010] CR2: 000055b3f7e14328 CR3: 00000001b5770000 CR4: 0000000000350ef0\n[ +0.000010] Call Trace:\n[ +0.000006] \n[ +0.000007] ? show_regs+0x6a/0x80\n[ +0.000018] ? __warn+0xa5/0x1b0\n[ +0.000019] ? mmu_interval_notifier_remove+0x327/0x340\n[ +0.000018] ? report_bug+0x24a/0x290\n[ +0.000022] ? handle_bug+0x46/0x90\n[ +0.000015] ? exc_invalid_op+0x19/0x50\n[ +0.000016] ? asm_exc_invalid_op+0x1b/0x20\n[ +0.000017] ? kasan_save_stack+0x26/0x50\n[ +0.000017] ? mmu_interval_notifier_remove+0x23b/0x340\n[ +0.000019] ? mmu_interval_notifier_remove+0x327/0x340\n[ +0.000019] ? mmu_interval_notifier_remove+0x23b/0x340\n[ +0.000020] ? __pfx_mmu_interval_notifier_remove+0x10/0x10\n[ +0.000017] ? kasan_save_alloc_info+0x1e/0x30\n[ +0.000018] ? srso_return_thunk+0x5/0x5f\n[ +0.000014] ? __kasan_kmalloc+0xb1/0xc0\n[ +0.000018] ? srso_return_thunk+0x5/0x5f\n[ +0.000013] ? __kasan_check_read+0x11/0x20\n[ +0.000020] amdgpu_hmm_unregister+0x34/0x50 [amdgpu]\n[ +0.004695] amdgpu_gem_object_free+0x66/0xa0 [amdgpu]\n[ +0.004534] ? __pfx_amdgpu_gem_object_free+0x10/0x10 [amdgpu]\n[ +0.004291] ? do_syscall_64+0x5f/0xe0\n[ +0.000023] ? srso_return_thunk+0x5/0x5f\n[ +0.000017] drm_gem_object_free+0x3b/0x50 [drm]\n[ +0.000489] amdgpu_gem_userptr_ioctl+0x306/0x500 [amdgpu]\n[ +0.004295] ? __pfx_amdgpu_gem_userptr_ioctl+0x10/0x10 [amdgpu]\n[ +0.004270] ? srso_return_thunk+0x5/0x5f\n[ +0.000014] ? __this_cpu_preempt_check+0x13/0x20\n[ +0.000015] ? srso_return_thunk+0x5/0x5f\n[ +0.000013] ? sysvec_apic_timer_interrupt+0x57/0xc0\n[ +0.000020] ? srso_return_thunk+0x5/0x5f\n[ +0.000014] ? asm_sysvec_apic_timer_interrupt+0x1b/0x20\n[ +0.000022] ? drm_ioctl_kernel+0x17b/0x1f0 [drm]\n[ +0.000496] ? __pfx_amdgpu_gem_userptr_ioctl+0x10/0x10 [amdgpu]\n[ +0.004272] ? drm_ioctl_kernel+0x190/0x1f0 [drm]\n[ +0.000492] drm_ioctl_kernel+0x140/0x1f0 [drm]\n[ +0.000497] ? __pfx_amdgpu_gem_userptr_ioctl+0x10/0x10 [amdgpu]\n[ +0.004297] ? __pfx_drm_ioctl_kernel+0x10/0x10 [d\n---truncated---", "severity": [ { "type": "CVSS_V3", "score": "CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H" } ], "affected": [ { "package": { "ecosystem": "Raspberry Pi OS:12", "name": "linux" }, "versions": [ "1:6.1.47-1+rpt4" ], "ecosystem_specific": { "binaries": [ { "binary_name": "linux-image-6.1.0-rpi3-rpi-2712", "binary_version": "1:6.1.47-1+rpt4" }, { "binary_name": "linux-image-6.1.0-rpi3-rpi-v6", "binary_version": "1:6.1.47-1+rpt4" }, { "binary_name": "linux-image-6.1.0-rpi3-rpi-v7", "binary_version": "1:6.1.47-1+rpt4" }, { "binary_name": "linux-image-6.1.0-rpi3-rpi-v7l", "binary_version": "1:6.1.47-1+rpt4" }, { "binary_name": "linux-image-6.1.0-rpi3-rpi-v8", "binary_version": "1:6.1.47-1+rpt4" } ] }, "database_specific": { "per_version": { "1:6.1.47-1+rpt4": { "linux_commit": "655fc658a15ae7a6f37103754adb39ba52a9a14e", "upstream_version": "6.1.47", "assessment": "commit ancestry (strak rule)", "upstream_fixed": "6.1.132", "compiled_filter": false } } } } ], "references": [ { "type": "ADVISORY", "url": "https://www.cve.org/CVERecord?id=CVE-2024-26656" }, { "type": "FIX", "url": "https://git.kernel.org/stable/c/2e13f88e01ae7e28a7e831bf5c2409c4748e0a60" }, { "type": "FIX", "url": "https://git.kernel.org/stable/c/e87e08c94c9541b4e18c4c13f2f605935f512605" }, { "type": "FIX", "url": "https://git.kernel.org/stable/c/af054a5fb24a144f99895afce9519d709891894c" }, { "type": "FIX", "url": "https://git.kernel.org/stable/c/22f665ecfd1225afa1309ace623157d12bb9bb0c" }, { "type": "FIX", "url": "https://git.kernel.org/stable/c/22207fd5c80177b860279653d017474b2812af5e" } ], "database_specific": { "source": "https://git.kernel.org/pub/scm/linux/security/vulns.git", "upstream_modified": "2026-08-05T11:52:36Z", "severity": { "source": "kernel CNA", "cvss_v3_score": 7.8, "cvss_v3_severity": "High", "scenario": "AV:L - Triggered via `DRM_IOCTL_AMDGPU_GEM_USERPTR` on a local AMD GPU device node (`/dev/dri/renderD*` or `card*`); there is no remote or adjacent-network path to `amdgpu_gem_userptr_ioctl()`.\nAC:L - A single ioctl with a fully attacker-chosen `addr`/`size` (e.g. 0xffffffffffff0000 / 0x80000000) deterministically overflows the interval-tree range check and reaches the unbalanced `mmdrop()`; there is no race and no dependence on memory layout.\nPR:L - The ioctl is flagged `DRM_RENDER_ALLOW`, so `drm_ioctl_permit()` waives `DRM_AUTH` for render clients \u2014 any unprivileged local user or sandboxed app with an open render node (standard for desktop sessions, Android apps, and GPU-enabled containers) qualifies, with no capability check anywhere on the path.\nUI:N - The attacker performs the entire sequence itself by opening the render node and issuing the ioctl; no victim action or cooperation is required.\nS:U - The corruption stays within the kernel's own security authority \u2014 the calling process's `mm_struct` and mmu-notifier tree \u2014 with no crossing of a VM, IOMMU, or hypervisor boundary.\nC:H - The unbalanced `mmdrop()` frees the live `mm_struct` and its PGD while still in use; reclaiming those objects yields page-table control and therefore arbitrary kernel and cross-process memory disclosure.\nI:H - The same use-after-free plus the `interval_tree_remove()` of a never-inserted node give heap-spray-controlled write primitives \u2014 freeing an in-use PGD allows forging page-table entries for arbitrary memory modification and privilege escalation.\nA:H - The path hits a `WARN_ON` (immediate panic under `panic_on_warn`), corrupts the mm's notifier interval tree, and tears down the running process's page tables, reliably crashing the kernel." } } }