SoylentNews

Arthur T Knackerbracket has processed the following story:

Just like a classical computer has separate, yet interconnected, components that must work together, such as a memory chip and a CPU on a motherboard, a quantum computer will need to communicate quantum information between multiple processors.

Current architectures used to interconnect superconducting quantum processors are “point-to-point” in connectivity, meaning they require a series of transfers between network nodes, with compounding error rates.

On the way to overcoming these challenges, MIT researchers developed a new interconnect device that can support scalable, “all-to-all” communication, such that all superconducting quantum processors in a network can communication directly with each other.

They created a network of two quantum processors and used their interconnect to send microwave photons back and forth on demand in a user-defined direction. Photons are particles of light that can carry quantum information.

The device includes a superconducting wire, or waveguide, that shuttles photons between processors and can be routed as far as needed. The researchers can couple any number of modules to it, efficiently transmitting information between a scalable network of processors.

They used this interconnect to demonstrate remote entanglement, a type of correlation between quantum processors that are not physically connected. Remote entanglement is a key step toward developing a powerful, distributed network of many quantum processors.

“In the future, a quantum computer will probably need both local and nonlocal interconnects. Local interconnects are natural in arrays of superconducting qubits. Ours allows for more nonlocal connections. We can send photons at different frequencies, times, and in two propagation directions, which gives our network more flexibility and throughput,” says Aziza Almanakly, an electrical engineering and computer science graduate student in the Engineering Quantum Systems group of the Research Laboratory of Electronics (RLE) and lead author of a paper on the interconnect.

The researchers previously developed a quantum computing module, which enabled them to send information-carrying microwave photons in either direction along a waveguide.

taylorvich writes:

https://phys.org/news/2025-03-decades-quest-antibiotic-compounds.html

A team of chemists, biologists and microbiologists led by researchers in Arts & Sciences at Washington University in St. Louis has found a way to tweak an antimalarial drug and turn it into a potent antibiotic, part of a project more than 20 years in the making. Importantly, the new antibiotic should be largely impervious to the tricks that bacteria have evolved to become resistant to other drugs.

"Antibiotic resistance is one of the biggest problems in medicine," said Timothy Wencewicz, an associate professor of chemistry in Arts & Sciences. "This is just one step on a long journey to a new drug, but we proved that our concept worked."

The findings are published in ACS Infectious Diseases. The lead author of the study, John Georgiades, AB '24, is now a graduate student at Princeton University who took over the project while he was an undergraduate in Wencewicz's lab. Other co-authors include Joseph Jez, the Spencer T. Olin Professor in Biology; Christina Stallings, a professor of molecular microbiology at the School of Medicine; and Bruce Hathaway, a professor emeritus at Southeast Missouri State University.

A new approach to antibiotics is sorely needed because many common drugs are losing their punch, Wencewicz said. He points to Bactrim, a combination of the drugs sulfamethoxazole and trimethoprim. Often prescribed to treat ear infections and urinary tract infections, Bactrim blocks a bacteria's ability to produce folate, an important nutrient for fast-growing germs.

"It's been prescribed so often that resistance is now very common," Wencewicz said. "For a long time, people have been thinking about what's going to replace Bactrim and where we go from here."

Instead of creating new antibiotics out of whole cloth, Georgiades, Wencewicz and their team used chemistry to tweak cycloguanil, an existing drug used to treat malaria. "It's a slick way to give new life to a drug that is already FDA-approved," Wencewicz said.

Like Bactrim, cycloguanil works by blocking the enzymes that organisms need to produce folate. It has saved millions of people from malaria over the decades, but it was useless against bacteria because it didn't have a way to penetrate the membrane that surrounds bacterial cells.

After many trials, researchers were able to attach various chemical keys to cycloguanil that opened the door to the bacterial membrane. Once the new compounds reached the inner workings of the cell, they staged a two-pronged attack on the enzymes that bacteria need to produce folate.

"Dual-action antibiotics tend to be much more effective than drugs that just take one approach," Wencewicz said. Bacteria may be able to evolve resistance to one part of the attack, but they won't easily find a way to stop both at once, he explained.

The new compound proved to be effective against a wide range of bacteria, including Escherichia coli and Staphylococcus aureus, two of the most common causes of bacterial infections. Unlike Bactrim and other existing drugs that target folate, some of the new compounds also showed power against Pseudomonas aeruginosa, a pathogen that often infects people with weakened immune systems.

More information: John D. Georgiades et al, Expanding the Landscape of Dual Action Antifolate Antibacterials through 2,4-Diamino-1,6-dihydro-1,3,5-triazines, ACS Infectious Diseases (2025). DOI: 10.1021/acsinfecdis.4c00768

Original Submission

taylorvich writes:

https://spectrum.ieee.org/jumping-robot

When you see a squirrel jump to a branch, you might think (and I myself thought, up until just now) that they're doing what birds and primates would do to stick the landing: just grabbing the branch and hanging on. But it turns out that squirrels, being squirrels, don't actually have prehensile hands or feet, meaning that they can't grasp things with any significant amount of strength. Instead, they manage to land on branches using a "palmar" grasp, which isn't really a grasp at all, in the sense that there's not much grabbing going on. It's more accurate to say that the squirrel is mostly landing on its palms and then balancing, which is very impressive.

This kind of dynamic stability is a trait that squirrels share with one of our favorite robots: Salto. Salto is a jumper too, and it's about as non-prehensile as it's possible to get, having just one limb with basically no grip strength at all. The robot is great at bouncing around on the ground, but if it could move vertically, that's an entire new mobility dimension that could lead to some potentially interesting applications, including environmental scouting, search and rescue, and disaster relief.

In a paper published today in Science Robotics, roboticists have now taught Salto to leap from one branch to another like squirrels do, using a low torque gripper and relying on its balancing skills instead.

While we're going to be mostly talking about robots here (because that's what we do), there's an entire paper by many of the same robotics researchers that was published in late February in the Journal of Experimental Biology about how squirrels land on branches this way. While you'd think that the researchers might have found some domesticated squirrels for this, they actually spent about a month bribing wild squirrels on the UC Berkeley campus to bounce around some instrumented perches while high speed cameras were rolling.

fliptop writes:

US sperm donor giant California Cryobank is warning customers it suffered a data breach that exposed customers' personal information:

California Cryobank is a full-service sperm bank providing frozen donor sperm and specialized reproductive services, such as egg and embryo storage. The company is the largest sperm bank in the US and services all 50 states and more than 30 countries worldwide.

California Cryobank detected suspicious activity on its network on April 21, 2024, and isolated the computers from the IT network.

"Through our investigation, CCB determined that an unauthorized party gained access to our IT environment and may have accessed and/or acquired files maintained on certain computer systems between April 20, 2024 and April 22, 2024," reads a from California Cryobank.

[...] An almost a year-long investigation has determined that the attack exposed varying personal data for customers, including names, bank accounts and routing numbers, Social Security numbers, driver's license numbers, payment card numbers, and/or health insurance information.

Original Submission

taylorvich writes:

https://newatlas.com/materials/carbon-negative-cement-sand-substitute-seawater-electricity-co2/

Concrete is the most widely used artificial material on the planet – which is a shame, because making it also happens to be one of the most polluting processes. Worse still, at a global scale it requires huge amounts of sand, which is getting harder (financially and environmentally) to mine from coasts, seafloors and riverbeds.

An unassuming new material from Northwestern could help solve both problems. Composed of calcium carbonate and magnesium hydroxide in different ratios, it's pretty simple to make – just take some seawater, zap it with electricity and bubble some CO2 through it.

The whole process is similar to how corals and mollusks build their shells, according to the team.

If you really want to get your thinking cap on, here's how they do it: two electrodes in the tank emit a low electrical current that splits the water molecules into hydrogen gas and hydroxide ions. When CO2 gas is added, the chemical composition of the water changes, increasing levels of bicarbonate ions. These hydroxide and bicarbonate ions react with other natural ions in seawater, producing solid minerals that gather at the electrodes.

The end result is a versatile white material that not only stores carbon, but can stand in for sand or gravel in cement, and also forms a foundational powder for other building materials like plaster and paint.

Intriguingly, the researchers found the material could be tweaked by adjusting the flow rate, timing and duration of the CO2 and seawater, and the voltage and current of the electricity. By tweaking the manufacturing process, the researchers can make the material with different properties for different purposes

Arthur T Knackerbracket has processed the following story:

While hunting in West Texas, a deer hunter spotted a strange object in a creek bed. Suspecting it might be a fossil, he took a photo and showed it to a ranch manager.

“I was skeptical,” O2 Ranch manager Will Juett said in a Sul Ross State University statement. “I figured it was likely just an old stump, but imagined how great it would be if he was right.”

The deer hunter was right, and the discovery was more than great, because it wasn’t just any fossil. An interdisciplinary team of researchers identified it as a mammoth tusk, an incredibly rare find for West Texas.

[...] “A local who subsequently wrote his PhD dissertation on it found one [a mammoth tusk] in Fort Stockton in the 1960s,” Schroeder said, adding that the specimen is currently the only mammoth tusk in Texas’ Trans-Pecos region to have been carbon-dated. “There was a big range of error [in carbon dating] back then. Now we can get it down to a narrower range within 500 years.”

While the statement doesn’t name a specific mammoth species, the tusk might have belonged to a Columbian mammoth, a distant cousin of the more familiar woolly mammoth. The shaggy elephantine animal could reach up to 13 feet in height (almost 4 meters) and weigh around 10 tons.

Columbian mammoths inhabited regions of North America, including modern-day Texas, before going extinct around 11,700 years ago along with many other Ice Age mammals. Though the reason behind the disappearance of the Ice Age’s iconic megafauna remains a hotly debated topic, scientists frequently cite climate change, and human hunting may have also played a role.

“Seeing that mammoth tusk just brings the ancient world to life,” Juett said. “Now, I can’t help but imagine that huge animal wandering around the hills on the O2 Ranch. My next thought is always about the people that faced those huge tusks with only a stone tool in their hand!”

Original Submission

fliptop writes:

Gaia runs faster on Ryzen AI PCs, using the XDNA NPU and RDNA iGPU:

Running large language models (LLMs) on PCs locally is becoming increasingly popular worldwide. In response, AMD is introducing its own LLM application, Gaia, an open-source project for running local LLMs on any Windows machine.

Gaia is designed to run various LLM models on Windows PCs and features further performance optimizations for machines equipped with its Ryzen AI processors (including the Ryzen AI Max 395+). Gaia uses the open-source Lemonade SDK from ONNX TurnkeyML for LLM inference. Models can allegedly adapt for different purposes with Gaia, including summarization and complex reasoning tasks.

[...] MD's new open-source project works by providing LLM-specific tasks through the Lemonade SDK and serving them across multiple runtimes. Lemonade allegedly "exposes an LLM web service that communicates with the GAIA application...via an OpenAI compatible Rest API." Gaia itself acts as an AI-powered agent that retrieves and processes data. It also "vectorizes external content (e.g., GitHub, YouTube, text files) and stores it in a local vector index."

Also at Phoronix.

AMD Press Release and GitHub Repository.

Original Submission

owl writes:

https://arstechnica.com/gadgets/2025/03/italian-court-orders-google-to-block-iptv-pirate-sites-at-dns-level/

Italy is using its Piracy Shield law to go after Google, with a court ordering the Internet giant to immediately begin poisoning its public DNS servers. This is just the latest phase of a campaign that has also targeted Italian ISPs and other international firms like Cloudflare. The goal is aimed at preventing illegal football streams, but the effort has already caused collateral damage. Regardless, Italy's communication regulator praises the ruling and hopes to continue sticking it to international tech firms.

The Court of Milan issued this ruling in response to a complaint that Google failed to block pirate websites after they were identified by the national communication regulator, known as AGCOM. The court found that the sites in question were involved in the illegal streaming of Series A football matches, which has been a focus of anti-piracy crusaders in Italy for years. Since Google offers a public DNS service, it is subject to the site-blocking law.

Piracy Shield is often labeled as draconian by opponents because blocking content via DNS is messy. It blocks the entire domain, which has led to confusion when users rely on popular platforms to distribute pirated content. Just last year, Italian ISPs briefly blocked the entire Google Drive domain because someone, somewhere used it to share copyrighted material. This is often called DNS poisoning or spoofing in the context of online attacks, and the outcome is the same if it's being done under legal authority: a DNS record is altered to prevent someone typing a domain name from being routed to the correct IP address.

Original Submission

An Anonymous Coward writes:

https://www.bleepingcomputer.com/news/security/microsoft-trust-signing-service-abused-to-code-sign-malware/

Cybercriminals are abusing Microsoft's Trusted Signing platform to code-sign malware executables with short-lived three-day certificates.

Threat actors have long sought after code-signing certificates as they can be used to sign malware to appear like they are from a legitimate company.

Signed malware also has the advantage of potentially bypassing security filters that would normally block unsigned executables, or at least treat them with less suspicion.

The holy grail for threat actors is to obtain Extended Validation (EV) code-signing certificates, as they automatically gain increased trust from many cybersecurity programs due to the more rigorous verification process. Even more important, EV certificates are believed to gain a reputation boost in SmartScreen, helping to bypass alerts that would normally be displayed for unknown files.

However, EV code-singing certificates can be difficult to obtain, requiring them to be stolen from other companies or for threat actors to set up fake businesses and spend thousands of dollars to purchase one. Furthermore, once the certificate is used in a malware campaign, it is usually revoked, making it unusable for future attacks.

Recently, cybersecurity researchers have seen threat actors utilizing the Microsoft Trusted Signing service to sign their malware with short-lived, three-day code-signing certificates.

[...] The Microsoft Trusted Signing service launched in 2024 and is a cloud-based service that allows developers to easily have their programs signed by Microsoft.

[...] "The service supports both public and private trust signing scenarios and includes a timestamping service."

[...] This increased security is accomplished by using short-lived certificates that can easily be revoked in the event of abuse and by never issuing the certificates directly to the developers, preventing them from being stolen in the event of a breach.

Arthur T Knackerbracket has processed the following story:

Over the past couple of weeks, I’ve been following news of the deaths of actor Gene Hackman and his wife, pianist Betsy Arakawa. It was heartbreaking to hear how Arakawa appeared to have died from a rare infection days before her husband, who had advanced Alzheimer’s disease and may have struggled to understand what had happened.

But as I watched the medical examiner reveal details of the couple's health, I couldn't help feeling a little uncomfortable. Media reports claim that the couple liked their privacy and had been out of the spotlight for decades. But here I was, on the other side of the Atlantic Ocean, being told what pills Arakawa had in her medicine cabinet, and that Hackman had undergone multiple surgeries.

It made me wonder: Should autopsy reports be kept private? A person’s cause of death is public information. But what about other intimate health details that might be revealed in a postmortem examination?

[...] The goal of an autopsy is to discover the cause of a person's death. Autopsy reports, especially those resulting from detailed investigations, often reveal health conditions—conditions that might have been kept private while the person was alive. There are multiple federal and state laws designed to protect individuals' health information. For example, the Health Insurance Portability and Accountability Act (HIPAA) protects "individually identifiable health information" up to 50 years after a person's death. But some things change when a person dies.

For a start, the cause of death will end up on the death certificate. That is public information. The public nature of causes of death is taken for granted these days, says Lauren Solberg, a bioethicist at the University of Florida College of Medicine. It has become a public health statistic. She and her student Brooke Ortiz, who have been researching this topic, are more concerned about other aspects of autopsy results.

The thing is, autopsies can sometimes reveal more than what a person died from. They can also pick up what are known as incidental findings. An examiner might find that a person who died following a covid-19 infection also had another condition. Perhaps that condition was undiagnosed. Maybe it was asymptomatic. That finding wouldn't appear on a death certificate. So who should have access to it?

taylorvich writes:

https://medicalxpress.com/news/2025-03-drug-reestablishes-brain-mouse.html

A new study by UCLA Health has discovered what researchers say is the first drug to fully reproduce the effects of physical stroke rehabilitation in model mice.

The findings, published in Nature Communications, tested two candidate drugs derived from their studies on the mechanism of the brain effects of rehabilitation, one of which resulted in significant recovery in movement control after stroke in mice.

Stroke is the leading cause of adult disability because most patients do not fully recover from the effects of stroke. There are no drugs in the field of stroke recovery, requiring stroke patients to undergo physical rehabilitation, which has shown to be only modestly effective.

"The goal is to have a medicine that stroke patients can take that produces the effects of rehabilitation," said Dr. S. Thomas Carmichael, the study's lead author and professor and chair of UCLA Neurology.

"Rehabilitation after stroke is limited in its actual effects because most patients cannot sustain the rehab intensity needed for stroke recovery.

"Further, stroke recovery is not like most other fields of medicine, where drugs are available that treat the disease—such as cardiology, infectious disease or cancer," Carmichael said.

"Rehabilitation is a physical medicine approach that has been around for decades; we need to move rehabilitation into an era of molecular medicine."

In the study, Carmichael and his team sought to determine how physical rehabilitation improved brain function after a stroke and whether they could generate a drug that could produce these same effects.

Working in laboratory mouse models of stroke and with stroke patients, the UCLA researchers identified a loss of brain connections that stroke produces that are remote from the site of the stroke damage. Brain cells located at a distance from the stroke site get disconnected from other neurons. As a result, brain networks do not fire together for things like movement and gait.

Arthur T Knackerbracket has processed the following story:

[Ed's Comment: Originally this story was viewable on FireFox and it downloaded fine using "Arthur". It is now giving a cookie warning that it cannot ever complete redirections and no longer displays. If anyone finds a solution to the problem please leave it in the comments. TY --JR]

There are a lot of things in life that we keep safely tucked away that we hope we'll never need to use. Our smoke alarms, for instance, or our emergency funds. These are the very things that we can't neglect, though, because when we need them, we really, really need them. Another solid example for drivers is a spare tire. Are you one of those unfortunate souls who has been stuck on an unfamiliar road late at night while waiting for your mechanic to hook you up with a spare? This topic is sure to strike a real chord with you, then. 

In November 2023, the UK's RAC reported that it had reviewed "equipment lists of more than 300 car models across 28 brands — everything from the smallest superminis to the largest 4x4s," and what did the British auto servicing brand discover? Less than 3% of those models were sold new with a spare wheel included in the price. 

For the manufacturer, of course, there's a money-saving benefit to limiting production of spares, while there are also some performance-related reasons to dispense with them. They add weight when kept in the back, and because they aren't always offered as full-size spares, they can limit performance while being driven on. As they're something of a last resort, drivers may not be inclined to use them anyway, which also limits the call for them. There are also more lightweight and convenient approaches to dealing with a flat, which is a further factor in the reduction of spare tires. 

If you were a fan of the fearsome muscle cars of the mid-to-late twentieth century, you surely still lament the fact that these mighty models became increasingly less practical, and then all but impossible to drive as a result of such paradigm shifts as the Clean Air Act. Enacted in 1970, the EPA reports that "this legislation authorized the development of comprehensive federal and state regulations to limit emissions from both stationary (industrial) sources and mobile sources," and there weren't many mobile sources more majestic than the Dodge Charger R/T (pictured here) and its kind. Fuel increasingly had to be cleaner, engines needed to be more efficient and generally smaller, and the trend for lighter, more practical models began.

 As important as a spare tire can be, there's no getting around the fact that it can add considerable weight to a vehicle: 44 pounds (20 kg) or so depending on the type of vehicle. This complicates the matter of hitting eco-friendlier targets. This could be seen as an advantage of the trend away from spare tires, having a potential positive effect on a vehicle's fuel economy, but the benefits of this compared to the risks associated with driving without a spare tire are a matter for the individual driver to decide on. 

After all, spares can certainly be hefty and unwieldy to work with at the roadside. Another part of the reasoning is that lots of drivers don't use them, which means they're often dead weight. Additionally, the vehicle not only has to store the wheel itself, but also the means to actually use it should the need arise. The jack alone can be quite the bulky accessory.

Booga1 writes:

https://petapixel.com/2025/03/19/us-court-of-appeals-unanimously-denies-copyright-protection-for-ai-created-images/

A unanimous federal appeals court ruled that pictures generated solely by machines do not qualify for copyright protection.
"The Copyright Act of 1976 requires all eligible work to be authorized in the first instance by a human being," said the U.S. Court of Appeals for the District of Columbia.
The 3-0 court ruling, issued March 18, was written by Circuit Judge Patricia A. Millett, who was nominated by President Obama in 2013.

Background
Computer scientist Dr. Stephen Thaler created a generative artificial intelligence named "Creativity Machine," which made a picture that Thaler titled "A Recent Entrance to Paradise."
The U.S. Copyright Office denied Thaler's application (for copyright registration) based on its requirement that work must be authored in the first instance by a human being. The copyright application listed Creativity Machine as the work's sole author.
Thaler litigated. A federal court (U.S. District Court for the District of Columbia) upheld the Copyright Office's denial; the federal appeals court affirmed the ruling of the federal district court.

After the March 18 opinion from the federal appeals court, Thaler's attorney, Ryan Abbott, said he and his client "strongly disagree" with the ruling and intend to appeal. The Copyright Office said it "believes the court reached the correct result."

"Judge Millett explained it best that, 'machines are tools, not authors.' Interpretations of the Copyright Act would be nonsensical if the 'author' could be a computer or other machine. Machines do not have children, they do not die, they do not have nationalities or hold property. All of these concepts referenced in copyright law would have absurd results if authorship was granted to a computer program, and courts are simply not allowed to re-interpret statutes or ignore portions of a statute." -- Alicia Calzada, Deputy General Counsel of the National Press Photographers Association (NPPA)

Previously: https://soylentnews.org/article.pl?sid=23/08/24/0036210

Original Submission

owl writes:

https://arstechnica.com/ai/2025/03/cloudflare-turns-ai-against-itself-with-endless-maze-of-irrelevant-facts/

On Wednesday, web infrastructure provider Cloudflare announced a new feature called "AI Labyrinth" that aims to combat unauthorized AI data scraping by serving fake AI-generated content to bots. The tool will attempt to thwart AI companies that crawl websites without permission to collect training data for large language models that power AI assistants like ChatGPT.

Cloudflare, founded in 2009, is probably best known as a company that provides infrastructure and security services for websites, particularly protection against distributed denial-of-service (DDoS) attacks and other malicious traffic.

Instead of simply blocking bots, Cloudflare's new system lures them into a "maze" of realistic-looking but irrelevant pages, wasting the crawler's computing resources. The approach is a notable shift from the standard block-and-defend strategy used by most website protection services. Cloudflare says blocking bots sometimes backfires because it alerts the crawler's operators that they've been detected.

Original Submission

An Anonymous Coward writes:

https://www.phoronix.com/news/Linux-6.15-slab

Ahead of the upcoming Linux 6.15 kernel cycle a few early pull requests have already been sent in to Linus Torvalds in advance of the anticipated v6.14 release on Sunday. Among those early changes for Linux 6.15 are the SLAB allocator updates that include a fix for cache randomization with kvmalloc inadvertently being inadequate due to accidentally using the same randomization seed.

With the SLAB pull request ahead of the Linux 6.15 merge window, most notable besides a few minor improvements is improving the kmalloc cache randomization within the kvmalloc code.

Google engineers discovered that the CONFIG_RANDOM_KMALLOC_CACHES hardening feature wasn't properly being employed. CONFIG_RANDOM_KMALLOC_CACHES creates multiple copies of slab caches and makes kmalloc randomly pick one based on the code address in order to help fend off memory vulnerability exploits. But the problem was the same random seed always ended up being used with the current Linux kernel code. From the Google code comments:

        "...This is problematic because `__kmalloc_node` will use the return address as the seed to derive the *random* cache to use. Since all calls to `kvmalloc_node` will use the same seed when the size is large, the hardening is rendered completely pointless."

Gong Ruiqi of Huawei who worked out the solution to the issue explained:

        "That literally means all kmalloc invoked via kvmalloc would use the same seed for cache randomization (CONFIG_RANDOM_KMALLOC_CACHES), which makes this hardening non-functional.

        The root cause of this problem, IMHO, is that using RET_IP only cannot identify the actual allocation site in case of kmalloc being called inside non-inlined wrappers or helper functions. And I believe there could be similar cases in other functions. Nevertheless, I haven't thought of any good solution for this. So for now let's solve this specific case first.

        For __kvmalloc_node_noprof, replace __kmalloc_node_noprof and call __do_kmalloc_node directly instead, so that RET_IP can take the return address of kvmalloc and differentiate each kvmalloc invocation."

At least with these pending SLAB updates for the Linux 6.15 merge window, this issue will be resolved and presumably be likely back-ported to existing stable kernels to address this ineffective security hardening.

- https://www.phoronix.com/news/Linux-6.15-Likely-Features
- https://lore.kernel.org/lkml/2f7985a8-0460-42de-9af0-4f966b937695@suse.cz/
- https://github.com/google/security-research/blob/908d59b573960dc0b90adda6f16f7017aca08609/pocs/linux/kernelctf/CVE-2024-27397_mitigation/docs/exploit.md?plain=1#L259
- https://patchwork.kernel.org/project/linux-mm/patch/20250212081505.2025320-3-gongruiqi1@huawei.com/

Original Submission