Search: t**n*e**l s.a. c***t**c***a

Organizations are rushing to implement AI without fully grasping where its legal protections begin and end. The post The AI Governance Gap Is a Leadership Problem: Waiting Won’t Close It appeared first on SecurityWeek .

Adform, a major advertising platform, remained compromised for roughly 24 hours from late on July 26 through the evening of July 27 after being breached by unknown attackers. Few people outside the industry recognize the name, but Adform serves around 1.5 billion ad impressions every day across tens of thousands of websites. That means anyone visiting any site that runs Adform ads could have been targeted. The attackers weren’t trying to install malware. Instead, they ran a script in the victim’s browser that checked the clipboard every three seconds, and if it found a cryptocurrency wallet address had been copied, swapped it for the attackers’ own wallet address. So if someone had a site with the malicious ad open in one browser tab, and was making a crypto transaction in another tab or in a dedicated app, the funds could have ended up in the attackers’ pockets instead. Adform’s owners caught the attack and fixed the problem, but there’s no guarantee a similar incident won’t happen again — which is why every user should defend themselves against malvertising . Check out our tips at the end of this post. What we know about the attack on Adform There isn’t a lot of information out there, since the company’s official statement covers only what happened and when, without getting into the root cause of the incident. Independent research has dug up technical details about how ordinary users were targeted, but none of that explains how Adform itself was breached in the first place. What’s clear is that the attackers slipped their own code into the JavaScript that loaded on every site running Adform ads. Whenever an ad was about to display, the script would load from Adform’s server, pick the right ad, and show it — but the attackers had tacked on a set of malicious functions: monitoring the clipboard, sending data about the site where the encounter happened, and the victim’s IP address back to their own server, and swapping out Bitcoin, Ethereum, and Tron wallet addresses. All it took to make it work was having any site with Adform-served ads open in one browser tab. It didn’t matter what kind of site it was, what the ad looked like, or which advertiser it belonged to. The one thing that mattered was whether the site ran over HTTP or HTTPS. According to Adform, the attack couldn’t succeed on a site loaded over HTTPS since the connection to the attackers’ server was blocked in this case. The company hasn’t shared any information on how many users were affected, or how many sites still serve their content and ads over HTTP. Malicious ads are an everyday occurrence Unfortunately, dangerous online ads have become a systemic problem. And we’re not just talking about sketchy supplement ads or gambling promos — we mean ads that spread malware or lead to sites designed to steal payment details and other valuable data. Attackers have built out industrial-scale infrastructure to pull this off, and they use several different approaches. Hacking and compromising ad servers. Adform isn’t an isolated case: attackers have previously breached Revive ad servers , for one, and spread malware through ads on PornHub . Hijacking the ad accounts of legitimate, reputable brands . All it takes is stealing a password from someone in marketing. From there, the cybercriminals run ads posing as the company they hacked, pushing fake app updates, bogus promotions, and similar scams. In the worst cases — like the account breaches at adtech.de and adxpansion.com — attackers managed to run ads that redirected victims straight into automatic malware installs ( drive-by downloads ). Buying ads directly. That’s right — attackers simply set up their own advertiser accounts and run ads for their phishing sites and malware, just like any other business online. Since ads show up practically everywhere — on websites, in apps, and on social media — these threats can turn up in pretty much any context. And you’ll find variations of this threat on both computers and mobile de…

This research highlights a design compromise in the Windows kernel related to Supervisor Mode Access Prevention (SMAP). Specifically, it experimentally confirms that SMAP is effectively disabled during standard IOCTL dispatches, allowing a stack pivot into user-mode memory without triggering SMAP protections. This behavior stems from the Windows kernel not being originally built with SMAP in mind, and retrofitting it would require extensive changes. The finding is a confirmation of an architectural limitation rather than a newly discovered vulnerability with an active exploit.

Marcus Hutchins doesn’t personally consider himself a hacker – but he accepts the epithet because it’s a widely used term for what he once did. The post Hacker Conversations: Marcus Hutchins and the Journey From the Gray Zone to Redemption appeared first on SecurityWeek .

Files built with the polyglot technique have been showing up more and more in cyberattacks in recent years. They let attackers slip malware past email filters and file scanners, deceive victims in phishing attacks, and complicate incident investigations. To pull this off, attackers deliberately construct a file that a system can interpret as different formats depending on which application opens it. A classic example is a file that can be handled as a PNG image or a ZIP archive. All it takes is changing the file’s extension, or simply using one or another application to open it. Let’s take a look at why it’s even possible to create files like this, which format combinations have turned up in real-world attacks, and how organizations can protect themselves from this threat. Why polyglot files are possible The data formats behind polyglot files are seldom exotic. It all comes down to a clever combination of common formats that happen to be structurally compatible. Polyglots exploit at least one of the following quirks in certain file formats: Most file formats need decoding from the very first byte, but some require to read them from the end. The clearest example is a ZIP archive: a corrupted or missing beginning doesn’t stop applications from reading the file, because all the needed headers actually sit at the end. This lets attackers simply glue two files together — in the example above, a PNG and a ZIP. The beginning reads as a valid PNG image, while the end reads as a valid ZIP archive. Many formats work like Russian matryoshka nesting dolls: despite outwardly having a specific extension matching its intended usage, inside the file is essentially a ZIP archive containing the necessary data. This group includes modern office documents (DOCX/XLSX/PPTX), Android installation packages (APK), Java library files (JAR), and many others. Some formats have no strict structural requirements — or their requirements are loose enough that the application that’s processing the file can locate the fragment it needs even when that fragment isn’t at the beginning. The Polydet GitHub repository describes numerous examples of possible file combinations for building a polyglot. Under the MITRE classification, this technique falls under the Masquerading category ( T1036.008 , Masquerade File Type). Examples of polyglot files in known cyberattacks Publicly available malware campaign analyses reveal all kinds of polyglots. Attackers adapt the entire attack scenario to fit a specific file type combination. The Head Mare group delivered the PhantomPyramid malware as a ZIP attachment. The file consisted of Windows executable code (EXE) with a small ZIP archive glued onto the end. When the victim opened the archive, it contained a file with the extension PDF.LNK, which would then launch that same polyglot attachment — this time as an executable file. In the attack documented by JPCERT , attackers created a file that started as a PDF and was detected as a PDF by most scanners but carried a DOC extension and opened in Office applications as a valid DOC file containing malicious macros. The attacks spreading the StrRAT and Ratty Trojans used a polyglot built from a signed Windows installer package (MSI) with malicious Java code (JAR) tacked onto the end. The StrelaStealer attacks used a polyglot with an HTML extension: a Windows library (DLL) with a decoy HTML document glued onto the end. A shortcut in the archive launched the file twice: once via the start command (the equivalent of a double-click, which opened a browser showing the HTML document), and once via rundll32 (which launched the malicious DLL). In a simulated but elegant attack, researchers joined two ordinary ZIP files together and found that different popular archive tools displayed the combined file differently: some showed only the first archive, some only the second, and some showed both at once as if it were a single archive with shared contents. If the attacker is familiar with the vic…

In the Linux kernel, the following vulnerability has been resolved: phonet: pep: fix use-after-free in pep_get_sb() pep_get_sb() doesn't consider that pskb_may_pull() might have relocated the skb data, and continue to access the older pointer, causing UAF. Reproduced under KASAN: BUG: KASAN: slab-use-after-free in pep_get_sb+0x234/0x3b0 Read of size 1 at addr ff11000105510f50 by task repro/157 pep_get_sb+0x234/0x3b0 pipe_handler_do_rcv+0x5f7/0xa10 pep_do_rcv+0x203/0x410 __sk_receive_skb+0x471/0x4a0 phonet_rcv+0x5b3/0x6c0 __netif_receive_skb+0xcc/0x1d0 Refetch the header with skb_header_pointer() after pskb_may_pull(), so the possibly stale pointer is no longer dereferenced. There are better ways to solve this, but, this is the less instrusive one.

In the Linux kernel, the following vulnerability has been resolved: comedi: comedi_parport: deal with premature interrupt Syzbot reported a general protection fault in `comedi_get_is_subdevice_running()`, which was called from the interrupt handler `parport_interrupt()` in the "comedi_parport" driver, but it does not currently have a C reproducer for the problem. It's probably due to a premature interrupt for one of two reasons: 1. The driver sets up the interrupt handler before the comedi subdevices used by the interrupt handler have been allocated, but does not disable the interrupt in the parallel port's CTRL register first. 2. The driver uses a user-supplied I/O port base address which Syzbot would have supplied, but it might not be backed by real parallel port hardware. Change the initialization order in the driver's comedi "attach" handler (`parport_attach()`) so that the hardware registers are initialized before the interrupt handler is requested. This should prevent premature interrupts occurring for real hardware. Also add a test to the interrupt handler to ensure the comedi device is fully attached and return early if it isn't.

In the Linux kernel, the following vulnerability has been resolved: exec: fix unsigned loop counter wrap in transfer_args_to_stack() The stop value is derived from bprm->p >> PAGE_SHIFT. The index variable is an unsigned long. If bprm->p drops below PAGE_SIZE and stop becomes zero the loop condition index >= stop is always true. After the index == 0 iteration the decrement wraps to ULONG_MAX and bprm->page[ULONG_MAX] reads sizeof(void *) bytes in front of the array. The pointer has wrapped to -1. That garbage pointer is then passed to kmap_local_page() and PAGE_SIZE bytes are copied from wherever that lands into the stack of the process being created. And the loop doesn't terminate either... Getting there only requires bprm->p < PAGE_SIZE. On !MMU bprm_set_stack_limit() and bprm_hit_stack_limit() are empty. So the only constraint on how far bprm->p is pushed down is valid_arg_len(), i.e. that each individual string still fits in what is left. bprm->p starts at PAGE_SIZE * MAX_ARG_PAGES - sizeof(void *) so a single argument or environment string of a little over 31 pages leaves it in the first page: Oops - load access fault [#1] CPU: 0 UID: 0 PID: 1 Comm: victim Not tainted 7.2.0-rc4 #1 epc : __memcpy+0xd4/0xf8 ra : transfer_args_to_stack+0xaa/0xae s4 : ffffffffffffffff s2 : 0000000000000000 a1 : ffffffdc98000000 a2 : 0000000000001000 status: 0000000a00001880 badaddr: ffffffdc98000000 cause: 0000000000000005 [<801a5324>] __memcpy+0xd4/0xf8 [<800d5f6a>] load_flat_binary+0x43a/0x65e [<800a2de4>] bprm_execve+0x1d4/0x316 [<800a351a>] do_execveat_common+0x12e/0x138 [<800a3d44>] __riscv_sys_execve+0x38/0x4e Kernel panic - not syncing: Fatal exception in interrupt This is an arcane bug but we should still fix it. Count down from MAX_ARG_PAGES so the loop ends when index reaches stop, stop == 0 included. The iterations performed are unchanged for every other value of stop. Only CONFIG_MMU=n builds are affected, transfer_args_to_stack() is used by binfmt_flat and binfmt_elf_fdpic on nommu only. The loop predates git history. commit 7e7ec6a93434 ("elf_fdpic_transfer_args_to_stack(): make it generic") only moved it from binfmt_elf_fdpic.c into fs/exec.c and narrowed the copy to the used part of the first page. The condition and the decrement are unchanged from 2.6.12-rc2.

In the Linux kernel, the following vulnerability has been resolved: drm/amdkfd: Check bounds on CRIU restore queue type and mqd size We weren't checking whether the values provided in the private data in kfd CRIU restore were within bounds. For queue type, add a KFD_QUEUE_TYPE_MAX and ensure the provided type is less than it. For mqd_size, add new function mqd_size_from_queue_type and confirm that the provided mqd_size matches expectations. (cherry picked from commit f19d8086f6644083c913d70bfdeee20e1b6f46a5)

In the Linux kernel, the following vulnerability has been resolved: drm/amdkfd: Check bounds in allocate_event_notification_slot The valid event ids go from 0 to KFD_SIGNAL_EVENT_LIMIT allocate_event_notification_slot has an option to specify an event id to allocate at, used by CRIU. We weren't checking the bounds on that value. Check them. v2: Lower bounds check is unecessary because of idr_alloc already rejecting negative numbers. Upper bounds check should be KFD_SIGNAL_EVENT_LIMIT since the signal mode mappings might not yet exist (cherry picked from commit 6853f1f6cbbeb3f53ebbbd7286536aeb2c5d5f50)

In the Linux kernel, the following vulnerability has been resolved: gpu: host1x: Fix use-after-free in host1x_bo_clear_cached_mappings __host1x_bo_unpin() drops the last reference to the mapping and frees it, so we can't dereference mapping afterwards. The cache itself outlives the mapping, so use the cache local variable instead.

In the Linux kernel, the following vulnerability has been resolved: fscrypt: Avoid dynamic allocation in fscrypt_get_devices() When a blk_crypto_key starts being used or is evicted, fs/crypto/ calls fscrypt_get_devices() to get the filesystem's list of block devices, then iterates over them and calls blk_crypto_config_supported(), blk_crypto_start_using_key(), or blk_crypto_evict_key() on each one. Currently, the block device pointers are placed in a dynamically allocated array. This dynamic allocation is problematic because: - It can fail, especially at the fscrypt_destroy_inline_crypt_key() call site when it's invoked for inode eviction under direct reclaim. - fscrypt_destroy_inline_crypt_key() doesn't handle the failure. It just zeroizes and frees the blk_crypto_key without calling blk_crypto_evict_key(). That causes a use-after-free. For now, let's fix this in the straightforward and easily-backportable way by switching to an on-stack array. Currently the fscrypt multi-device functionality is used only by f2fs, which has a hardcoded limit of 8 block devices. An on-stack array works fine for that. (Of course, this solution won't scale up to large number of block devices. For that we'd need a different solution, like moving the block device iteration into the filesystem. Or in the case of btrfs, which will only support blk-crypto-fallback, we should make it just call blk-crypto-fallback directly, so the block devices won't be needed.)

In the Linux kernel, the following vulnerability has been resolved: fscrypt: Add missing superblock check in find_or_insert_direct_key() The legacy 'fscrypt_direct_keys' table caches master keys that are used by v1 encryption policies that have FSCRYPT_POLICY_FLAG_DIRECT_KEY. It's just a global table for all filesystems (since the keys can be provided by the legacy process-subscribed keyrings mechanism, which makes it difficult to reuse super_block::s_master_keys). The entries in it ('struct fscrypt_direct_key') do contain a super_block pointer, though, for passing to fscrypt_destroy_inline_crypt_key() when the last inode that references the key is evicted. However, when finding the fscrypt_direct_key for an inode, we weren't actually comparing the super_block pointer. As a result, inodes with different super_blocks could point to the same fscrypt_direct_key. That could extend the lifetime of a fscrypt_direct_key beyond the super_block it points to, causing a use-after-free later. Fix this by creating distinct fscrypt_direct_key structs for distinct super_block structs. Note that this problem doesn't exist in the v2 policy equivalent ("per-mode keys"), since the data structures there are per super_block.

In the Linux kernel, the following vulnerability has been resolved: mm/damon/core: disallow overlapping input ranges for damon_set_regions() damon_set_regions() assumes the input ranges are sorted by the address and don't overlap each other. Hence the assumption was initially to be explicitly validated. But commit 97d482f4592f ("mm/damon/sysfs: reuse damon_set_regions() for regions setting") has mistakenly removed the validation. This can make DAMON behave in unexpected ways. At the best, the monitoring results snapshot will just look weird since there will be overlapping regions. DAMOS will also work weirdly, applying the same action multiple times for overlapping regions, and make DAMOS quota weird. More seriously, depending on the setup and regions updates sequence, negative size regions can be made. It will trigger WARN_ONCE() if the kernel is built with CONFIG_DAMON_DEBUG_SANITY=y. Depending on the monitoring results, the negative size region can further trigger division by zero in damon_merge_two_regions(). Note that some of the consequences including the WARN_ONCE() and the divide by zero depend on commits that were introduced after the root cause commit 97d482f4592f ("mm/damon/sysfs: reuse damon_set_regions() for regions setting"). Fix the problems by checking the assumption and returning an error if the input ranges don't meet the assumption. The issue was discovered [1] by Sashiko.

In the Linux kernel, the following vulnerability has been resolved: mm/damon/core: validate ranges in damon_set_regions() DAMON core logic assumes zero length regions don't exist. However, a few DAMON API callers including DAMON_SYSFS, DAMON_RECLAIM and DAMON_LRU_SORT allow users to set empty monitoring target regions. This could result in WARN_ONCE() on CONFIG_DAMON_DEBUG_SANITY enabled kernel, and divide-by-zero from damon_merge_two_regions(). For example, the WANR_ONCE() can be triggered like below. # grep DAMON_DEBUG_SANITY /boot/config-$(uname -r) # CONFIG_DAMON_DEBUG_SANITY=y # damo start # cd /sys/kernel/mm/damon/admin/kdamonds/0 # echo 0 > contexts/0/targets/0/regions/0/start # echo 0 > contexts/0/targets/0/regions/0/end # echo commit > state # dmesg [....] [ 73.705780] ------------[ cut here ]------------ [ 73.707552] start 0 >= end 0 [ 73.708452] WARNING: mm/damon/core.c:359 at damon_new_region+0x6e/0x80, CPU#1: kdamond.0/758 [...] All DAMON API callers eventually use damon_set_regions() to setup the regions. Add the validation logic in the function.

In the Linux kernel, the following vulnerability has been resolved: userfaultfd: prevent registration of special VMAs Vova Tokarev says: userfaultfd allows registration on shadow stack VMAs. With userfaultfd access, you can register on the shadow stack, discard a page ... and inject a page with chosen return addresses via UFFDIO_COPY. Update vma_can_userfault() to reject VM_SHADOW_STACK. While on it, also reject VM_SPECIAL so that if a driver would implement vm_uffd_ops, it wouldn't be possible to register special VMAs with userfaultfd. Since VM_SPECIAL includes VM_DONTEXPAND which is set but hugetlb, exclude hugetlb VMAs from the check for VM_SPECIAL.

In the Linux kernel, the following vulnerability has been resolved: arm64: make huge_ptep_get handled unaligned addresses huge_ptep_get() can be handed a virtual address pointing to the middle of a contpmd/contpte mapped hugetlb folio (examples of callers are pagemap_hugetlb_range, page_mapped_in_vma). The arm64 helper rewalks the pgtables in find_num_contig to answer whether the huge pte we have maps a contpmd or a contpte hugetlb folio, and returns CONT_PMDS or CONT_PTES, so that it can collect a/d bits over the contiguous ptes. We can falsely return CONT_PTES instead of CONT_PMDS if the addr is not aligned. On systems where CONT_PTES != CONT_PMDS (meaning page size is 16K), we could collect excess A/D bit state, meaning extra work for the kernel. Even worse, we may iterate beyond the PTE table and dereference a garbage ptep pointer to access physical memory we don't own. Since the ptep pointer is a linear map address, we may run off the end of the linear map or into a hole, dereference a VA not mapped into the kernel pgtables and cause kernel panic. Fix this by aligning the pmdp pointer down to a contpmd base before checking equality with the passed huge pte pointer, to correctly answer whether the huge pte is the base of a contpmd block.

In the Linux kernel, the following vulnerability has been resolved: ALSA: timer: don't re-enter an instance callback that is still running The userspace-driven timer (utimer) TRIGGER ioctl calls snd_timer_interrupt() directly with no serialization, so two threads triggering the same utimer can run snd_timer_interrupt() on one snd_timer concurrently. snd_timer_process_callbacks() drops timer->lock around each instance callback and marks the in-flight callback with the single SNDRV_TIMER_IFLG_CALLBACK bit; snd_timer_close_locked() waits on that bit to drain an in-flight callback before freeing the instance. The bit cannot represent two concurrent callbacks: when a second interrupt re-queues an instance whose callback is still running, both run at once, the first to finish clears the bit, and the close-path drain then frees the instance (and its callback_data) while the other callback is still live - a use-after-free reachable by any user able to open /dev/snd/timer, both via a user timer instance and via a sequencer queue timer bound to the utimer. snd_timer_interrupt() sets IFLG_CALLBACK before dropping timer->lock, so a concurrent interrupt already observes it under the lock. Skip re-queuing an instance (and its slaves) to the ack/sack list while its callback is in flight; the accumulated pticks are delivered on the next tick, so no event is lost.

In the Linux kernel, the following vulnerability has been resolved: ALSA: timer: drain a slave's callback before its master detaches it snd_timer_close_locked() drains the closing instance's own in-flight callback (IFLG_CALLBACK) before freeing it, but not its slaves'. When a master instance is closed, remove_slave_links() clears each slave's ->timer; the slave's own close then reads timer == NULL and takes the branch that skips the drain entirely (snd_timer_stop_slave() also no-ops on a NULL timer). So a slave whose callback is still running when the master is closed is freed underneath the live callback, leading to use-after-free. Drain the slaves too before remove_slave_links() severs them. snd_timer_stop() has already taken this instance off the active list, so no new slave callback can be queued. Take the slaves off the ack list so a pending one can't fire either, then wait for any that is already in flight.

In the Linux kernel, the following vulnerability has been resolved: drm/imagination: acquire vm_ctx->lock before mapping memory to GPU VM The drm gpuvm code doesn't protect find operation against map operation, and the driver needs to ensure a map operation shouldn't happen when a find operation is in progress. In some cases a find operation will be in progress when doing map/unmap operations, and the find operation will do a NULL pointer dereference. An example of the stack trace of such NULL dereference is shown below: ``` Unable to handle kernel access to user memory without uaccess routines at virtual address 0000000000000010 [<ffffffff01e989d4>] drm_gpuva_find+0x28/0x6c [drm_gpuvm] [<ffffffff01ed3a40>] pvr_vm_unmap+0x34/0x68 [powervr] [<ffffffff01ec69da>] pvr_ioctl_vm_unmap+0x2e/0x50 [powervr] [<ffffffff8080ce0a>] drm_ioctl_kernel+0x8e/0xdc [<ffffffff8080d016>] drm_ioctl+0x1be/0x3e0 [<ffffffff802bec3e>] __riscv_sys_ioctl+0xba/0xc4 [<ffffffff80d858b2>] do_trap_ecall_u+0x23e/0x3f4 [<ffffffff80d92288>] handle_exception+0x168/0x174 ``` As all occurences of drm_gpuva_find*() are already guarded by vm_ctx->lock, make pvr_vm_map() to acquire this lock to prevent disturbing any find operation. This fixes the NULL deference problem in drm_gpuva_find*().