Connect with us

FACEBOOK

‘Carol’s Journey’: What Facebook knew about how it radicalized users – NBC News

Published

on

In the summer of 2019, a new Facebook user named Carol Smith signed up for the platform, describing herself as a politically conservative mother from Wilmington, North Carolina. Smith’s account indicated an interest in politics, parenting, and Christianity, and followed a few of her favorite brands, including Fox News and then-President Donald Trump. 

Though Smith had never expressed interest in conspiracy theories, in just two days Facebook was recommending she join groups dedicated to QAnon, a sprawling and baseless conspiracy theory and movement that claimed Trump was secretly saving the world from a cabal of pedophiles and Satanists.

Smith didn’t follow the recommended QAnon groups, but whatever algorithm Facebook was using to determine how she should engage with the platform pushed ahead just the same. Within one week, Smith’s feed was full of groups and pages that had violated Facebook’s own rules, including those against hate speech and disinformation.

Smith wasn’t a real person. A researcher employed by Facebook invented the account, along with those of other fictitious “test users” in 2019 and 2020, as part of an experiment in studying the platform’s role in misinforming and polarizing users through its recommendations systems.

That researcher said Smith’s Facebook experience was “a barrage of extreme, conspiratorial, and graphic content.” 

Advertisement
free widgets for website

The body of research consistently found Facebook pushed some users into “rabbit holes,” increasingly narrow echo-chambers where violent conspiracy theories thrived. People radicalized through these rabbit holes make up a small slice of total users, but at Facebook’s scale, that can mean millions of individuals.

The findings, communicated in a report titled “Carol’s Journey to QAnon,” were among thousands of pages of documents included in disclosures made to the Securities and Exchange Commission and provided to Congress in redacted form by legal counsel for Frances Haugen, who worked as a Facebook product manager until May. Haugen is now asserting whistleblower status and has filed several specific complaints that Facebook puts profit over public safety. Earlier this month, she testified about her claims before a Senate subcommittee

Versions of the disclosures — which redacted the names of researchers, including the author of “Carol’s Journey to QAnon” — were shared digitally and reviewed by a consortium of news organizations, including NBC News. The Wall Street Journal published a series of reports based on many of the documents last month. 

“While this was a study of one hypothetical user, it is a perfect example of research the company does to improve our systems and helped inform our decision to remove QAnon from the platform,” a Facebook spokesperson said in a response to emailed questions.

Facebook CEO Mark Zuckerberg has broadly denied Haugen’s claims, defending his company’s “industry-leading research program” and its commitment “to identify important issues and work on them.” The documents released by Haugen partly support those claims, but also highlight the frustrations of some of the employees engaged in that research. 

Advertisement
free widgets for website

Among Haugen’s disclosures are research, reports and internal posts that suggest Facebook has long known that its algorithms and recommendation systems push some users to extremes. And while some managers and executives ignored the internal warnings, anti-vaccine groups, conspiracy theory movements and disinformation agents took advantage of their permissiveness, threatening public health, personal safety and democracy at large.  

“These documents effectively confirm what outside researchers were saying for years prior, which was often dismissed by Facebook,” said Renée DiResta, technical research manager at the Stanford Internet Observatory and one of the earliest harbingers of the risks of Facebook’s recommendation algorithms. 

Facebook’s own research shows how easily a relatively small group of users has been able to hijack the platform, and for DiResta, settles any remaining question about Facebook’s role in the growth of conspiracy networks. 

See also  Facebook's Own Research Says Its Apps Can Harm Mental Health. Senators Have Questions - NPR

“Facebook literally helped facilitate a cult,” she said. 

‘A pattern at Facebook’

Advertisement
free widgets for website

For years, company researchers had been running experiments like Carol Smith’s to gauge the platform’s hand in radicalizing users, according to the documents seen by NBC News.

This internal work repeatedly found that recommendation tools pushed users into extremist groups, a series of disclosures that helped inform policy changes and tweaks to recommendations and newsfeed rankings. Those rankings are a tentacled, ever-evolving system widely known as “the algorithm” that pushes content to users. But the research at that time stopped well short of inspiring any movement to change the groups and pages themselves.

That reluctance was indicative of “a pattern at Facebook,” Haugen told reporters this month. “They want the shortest path between their current policies and any action.”

Haugen added, “There is great hesitancy to proactively solve problems.” 

A Facebook spokesperson disputed that the research had not pushed the company to act and pointed to changes to groups announced in March.

Advertisement
free widgets for website

While QAnon followers committed real-world violence in 2019 and 2020, groups and pages related to the conspiracy theory skyrocketed, according to internal documents. The documents also show how teams inside Facebook took concrete steps to understand and address those issues — some of which employees saw as too little, too late.  

By the summer of 2020, Facebook was hosting thousands of private QAnon groups and pages, with millions of members and followers, according to an unreleased internal investigation

A year after the FBI designated QAnon as a potential domestic terrorist threat in the wake of armed standoffs, kidnappings, harassment campaigns and shootings, Facebook labeled QAnon a “Violence Inciting Conspiracy Network,” and banned it from the platform, along with militias and other violent social movements. A small team working across several of Facebook’s departments had hosted hundreds of ads on Facebook and Instagram worth thousands of dollars and millions of views, “praising, supporting, or representing” the conspiracy theory.

The Facebook spokesperson said in an email that the company has “taken a more aggressive approach in how we reduce content that is likely to violate our policies, in addition to not recommending Groups, Pages or people that regularly post content that is likely to violate our policies.

For many employees inside Facebook, the enforcement came too late, according to posts left on Workplace, the company’s internal message board. 

Advertisement
free widgets for website

“We’ve known for over a year now that our recommendation systems can very quickly lead users down the path to conspiracy theories and groups,” one integrity researcher, whose name had been redacted, wrote in a post announcing she was leaving the company. “This fringe group has grown to national prominence, with QAnon congressional candidates and QAnon hashtags and groups trending in the mainstream. We were willing to act only * after * things had spiraled into a dire state.” 

‘We should be concerned’

While Facebook’s ban initially appeared effective, a problem remained. The removal of groups and pages didn’t wipe out QAnon’s most extreme followers, who continued to organize on the platform.

“There was enough evidence to raise red flags in the expert community that Facebook and other platforms failed to address QAnon’s violent extremist dimension,” said Marc-André Argentino, a research fellow at King’s College London’s International Centre for the Study of Radicalisation, who has extensively studied QAnon. 

Advertisement
free widgets for website

Believers simply rebranded as anti-child trafficking groups or migrated to other communities, including those around the anti-vaccine movement. 

It was a natural fit. Researchers inside Facebook studying the platform’s niche communities found violent conspiratorial beliefs to be connected to Covid vaccine hesitancy. In one study, researchers found QAnon community members were also highly concentrated in anti-vaccine communities. Anti-vaccine influencers had similarly embraced the opportunity of the pandemic, and used Facebook’s features like groups and livestreaming to grow their movements. 

See also  Newsmax's $25K Trump Facebook Ad Boosted Fossella's Staten Island Borough President Bid

“We do not know if QAnon created the preconditions for vaccine hesitancy beliefs,” researchers wrote. “It may not matter either way. We should be concerned about people affected by both problems.”

QAnon believers also jumped to groups promoting President Donald Trump’s false claim that the 2020 election was stolen, groups that trafficked in a hodgepodge of baseless conspiracy theories alleging voters, Democrats and election officials were somehow cheating Trump out of a second term. This new coalition, largely organized on Facebook, ultimately stormed the U.S. Capitol on Jan. 6, according to a report included in the document trove and first reported by Buzzfeed News in April. 

These conspiracy groups had become the fastest-growing groups on all of Facebook, according to the report, but Facebook wasn’t able to control their “meteoric growth,” the researchers wrote, “because we were looking at each entity individually, rather than as a cohesive movement.” A Facebook spokesperson told BuzzFeed News it took many steps to limit election misinformation but that it was unable to catch everything.

Advertisement
free widgets for website

Facebook’s enforcement was “piecemeal,” the team of researchers wrote, noting, “we’re building tools and protocols and having policy discussions to help us do this better next time.” 

‘A head-heavy problem’

The attack on the Capitol invited harsh self-reflection from employees. 

One team invoked the lessons learned during QAnon’s moment to warn about permissiveness with anti-vaccine groups and content, which researchers found comprised up to half of all vaccine content impressions on the platform. 

Advertisement
free widgets for website

“In rapidly-developing situations, we’ve often taken minimal action initially due to a combination of policy and product limitations making it extremely challenging to design, get approval for, and roll out new interventions quickly.” the report said. Qanon was offered as an example of an example of a time when Facebook was “prompted by societal outcry at the resulting harms to implement entity takedowns” for a crisis on which “we initially took limited or no action.” 

The effort to overturn the election also invigorated efforts to clean up the platform in a more proactive way. 

Facebook’s “Dangerous Content” team formed a working group in early 2021 to figure out ways to deal with the kind of users who had been a challenge for Facebook: communities including QAnon, Covid-denialists and the misogynist incel movement that weren’t obvious hate or terrorism groups, but that, by their nature, posed a risk to the safety of individuals and societies. 

The focus wasn’t to eradicate them, but to curb the growth of these newly branded “harmful topic communities,” with the same algorithmic tools that had allowed them to grow out of control. 

“We know how to detect and remove harmful content, adversarial actors, and malicious coordinated networks, but we have yet to understand the added harms associated with the formation of harmful communities, as well as how to deal with them,” the team wrote in a 2021 report.

In a February 2021 report, they got creative. An integrity team details an internal system meant to measure and protect users against societal harms including radicalization, polarization, and discrimination that its own recommendation systems had helped cause. Building on a previous research effort dubbed “Project Rabbithole,” the new program was dubbed Drebbel. Cornelis Drebbel was a 17th-century Dutch engineer known for inventing the first navigable submarine and the first thermostat. 

Advertisement
free widgets for website

The Drebbel group was tasked with discovering and ultimately stopping the paths that moved users towards harmful content on Facebook and Instagram, including in anti-vax and QAnon groups. A post from the Drebbel team praised the earlier research on test users. “We believe Drebbel will be able to scale this up significantly,” they wrote.

See also  Facebook Teases Technology That Allows Your Brain To Control A Computer

“Group joins can be an important signal and pathway for people going towards harmful

and disruptive communities,” the group stated in a post to Workplace, Facebook’s internal message board. “Disrupting this path can prevent further harm.”

The Drebbel group features prominently in Facebook’s “Deamplification Roadmap,” a multi-step plan published on the company Workplace on Jan. 6, that includes a complete audit of recommendation algorithms.

In March, the Drebbel group posted about their progress via a study and suggested a way forward. If researchers could systematically identify the “gateway groups,” those that fed into anti-vaccination and QAnon communities, they wrote, maybe Facebook could put up roadblocks to keep people from falling through the rabbit hole. 

Advertisement
free widgets for website

The Drebbel “Gateway Groups” study looked back at a collection of QAnon and anti-vaccine groups that had been removed for violating policies around misinformation and violence and incitement. It used the membership of these purged groups to study how users had been pulled in. Drebbel identified 5,931 QAnon groups with 2.2 million total members, half of which joined through so-called gateway groups. For 913 anti-vaccination groups with 1.7 million members, the study identified one million gateway groups (Facebook has said it recognizes the need to do more).

Facebook integrity employees warned in an earlier report that anti-vaccine groups could become more extreme. 

“Expect to see a bridge between online and offline world,” the report said. “We might see motivated users create sub-communities with other highly motivated users to plan action to stop vaccination.”

A separate cross-department group reported this year that vaccine hesitancy in the U.S. “closely resembled” QAnon and Stop the Steal movements, “primarily driven by authentic actors and community building.” 

“We found, like many problems at FB,” the team wrote, “that this is a head-heavy problem with a relatively few number of actors creating a large percentage of the content and growth.”

Advertisement
free widgets for website

The Facebook spokesperson said that the company had “focused on outcomes” in relation to Covid-19 and that it had seen vaccine hesitancy decline by 50 percent, according to a survey it conducted with Carnegie-Mellon University and University of Maryland.

Whether Facebook’s newest integrity initiatives will be able to stop the next dangerous conspiracy theory movement or the violent organization of existing movements remains to be seen. But their policy recommendations may carry more weight now that the violence on Jan. 6 laid bare the outsized influence and dangers of even the smallest extremist communities and the misinformation that fuels them. 

“The power of community, when based on harmful topics or ideologies, potentially poses a greater threat to our users than any single piece of content, adversarial actor, or malicious network,” a 2021 report concluded.

The Facebook spokesperson said that the recommendations in the “Deamplification Roadmap” are on track: “This is important work and we have a long track record of using our research to inform changes to our apps,” the spokesperson wrote. “Drebbel is consistent with this approach, and its research helped inform our decision this year to permanently stop recommending civic, political or news Groups on our platforms. We are proud of this work and we expect it to continue to inform product and policy decisions going forward.”

CORRECTION (Oct. 22, 2021, 7:06 p.m. ET): A previous version of this article misstated the status of groups studied by Facebook’s Drebbel team. It looked at groups that Facebook had removed, not those that were currently active.

Advertisement
free widgets for website

Read More

Continue Reading
Advertisement free widgets for website
Click to comment

Leave a Reply

Your email address will not be published.

FACEBOOK

Upcoming Restriction Period for US ads about social issues, elections, or politics

Published

on

By

upcoming-restriction-period-for-us-ads-about-social-issues,-elections,-or-politics

In recent years, Meta has developed a comprehensive approach to protecting elections on our technologies. These efforts continue in advance of the US 2022 Midterms, which you can read more about in our Newsroom.

Implementing a restriction period for ads about social issues, elections or politics in the US

Consistent with our approach during the US 2020 General Election, we are introducing a restriction period for ads about social issues, elections or politics in the US. The restriction period will run from 12:01 AM PT on Tuesday, November 1, 2022 through 11:59 PM PT on Tuesday, November 8, 2022.

We are putting this restriction period in place again because we found that the restriction period achieves the right balance of giving campaigns a voice while providing additional time for scrutiny of issue, electoral, and political ads in the Ad Library. We are sharing the requirements and key dates ahead of time, so advertisers are able to prepare their campaigns in the months and weeks ahead.

What to know about the ad restriction period in the US

We will not allow any new ads about social issues, elections or politics in the US from 12:01 AM PT on Tuesday, November 1, 2022 through 11:59 PM PT on Tuesday, November 8, 2022.

In order to run ads about social issues, elections or politics in the US during the restriction period, the ads must be created with a valid disclaimer and have delivered an impression prior to 12:01 AM PT on Tuesday, November 1, 2022, but with limited editing capabilities.

Advertisement
free widgets for website

What advertisers can do during the restriction period for eligible ads:

  • Edit bid amount, budget amount and scheduled end date
  • Pause and unpause eligible ads that have already served at least 1 impression with a valid disclaimer prior to the restriction period going into effect
See also  Facebook 'Supreme Court' Orders Social Network To Restore 4 Posts In First Rulings

What advertisers cannot do during the restriction period for eligible ads, includes but is not limited to:

  • Editing certain aspects of eligible ads, such as ad creative (including ad copy, image/video assets, website URL)
  • Editing targeting, placement, optimization or campaign objective
  • Removing or adding a disclaimer
  • Copy, duplicating or boosting ads

See the Help Center for detailed requirements of what is or isn’t allowed during the restriction period.

Planning ahead for key dates

Keep in mind the following dates as you plan your campaign to avoid delays or disapprovals that may prevent your ads from running during the restriction period:

  • By Tuesday, October 18, 2022: Complete the ad authorization process to get authorized to run ads about social issues, elections or politics, which includes setting up an approved disclaimer for your ads.

  • By Tuesday, October 25, 2022: Submit your issue, electoral or political ads in order to best ensure that your ads are live and have delivered at least 1 impression with a valid disclaimer before the restriction period begins.
    • Please ensure that you add your approved disclaimer to these ads by choosing ISSUES_ELECTIONS_POLITICS in the special_ad_categories field. You will not be able to add a disclaimer after 12:01 AM PT on Tuesday, November 1, 2022.

  • Between Tuesday, November 1, 2022 and Tuesday, November 8, 2022: The ad restriction period will be in effect. We will not allow any new ads to run about social issues, elections or politics in the US starting 12:01 AM PT on Tuesday, November 1 through 11:59 PM PT on Tuesday, November 8, 2022.
  • At 12:00 AM PT on Wednesday, November 9, 2022: We will allow new ads about social issues, elections or politics to be published.

As the restriction period approaches, we encourage you to review these ad restriction period best practices to properly prepare ahead of time.

We will continue to provide updates on our approach to elections integrity or on any changes regarding the restriction period via this blog.

Visit the Elections Hub or our FAQ for more advertising resources.

First seen at developers.facebook.com

Advertisement
free widgets for website
Continue Reading

FACEBOOK

Signals in prod: dangers and pitfalls

Published

on

By

signals-in-prod:-dangers-and-pitfalls

In this blog post, Chris Down, a Kernel Engineer at Meta, discusses the pitfalls of using Linux signals in Linux production environments and why developers should avoid using signals whenever possible.

What are Linux Signals?

A signal is an event that Linux systems generate in response to some condition. Signals can be sent by the kernel to a process, by a process to another process, or a process to itself. Upon receipt of a signal, a process may take action.

Signals are a core part of Unix-like operating environments and have existed since more or less the dawn of time. They are the plumbing for many of the core components of the operating system—core dumping, process life cycle management, etc.—and in general, they’ve held up pretty well in the fifty or so years that we have been using them. As such, when somebody suggests that using them for interprocess communication (IPC) is potentially dangerous, one might think these are the ramblings of someone desperate to invent the wheel. However, this article is intended to demonstrate cases where signals have been the cause of production issues and offer some potential mitigations and alternatives.

Signals may appear attractive due to their standardization, wide availability and the fact that they don’t require any additional dependencies outside of what the operating system provides. However, they can be difficult to use safely. Signals make a vast number of assumptions which one must be careful to validate to match their requirements, and if not, one must be careful to configure correctly. In reality, many applications, even widely known ones, do not do so, and may have hard-to-debug incidents in the future as a result.

Let us look into a recent incident that occurred in the Meta production environment, reinforcing the pitfalls of using signals. We’ll go briefly over the history of some signals and how they led us to where we are today, and then we’ll contrast that with our current needs and issues that we’re seeing in production.

Advertisement
free widgets for website

The Incident

First, let’s rewind a bit. The LogDevice team cleaned up their codebase, removing unused code and features. One of the features that was deprecated was a type of log that documents certain operations performed by the service. This feature eventually became redundant, had no consumers and as such was removed. You can see the change here on GitHub. So far, so good.

The next little while after the change passed without much to speak about, production continued ticking on steadily and serving traffic as usual. A few weeks later, a report that service nodes were being lost at a staggering rate was received. It was something to do with the rollout of the new release, but what exactly was wrong was unclear. What was different now that had caused things to fall over?

The team in question narrowed the problem to the code change we mentioned earlier, deprecating these logs. But why? What’s wrong with that code? If you don’t already know the answer, we invite you to look at that diff and try to work out what’s wrong because it’s not immediately obvious, and it’s a mistake anyone could make.

logrotate, Enter the Ring

logrotate is more or less the standard tool for log rotation when using Linux. It’s been around for almost thirty years now, and the concept is simple: manage the life cycle of logs by rotating and vacuuming them.

logrotate doesn’t send any signals by itself, so you won’t find much, if anything, about them in the logrotate main page or its documentation. However, logrotate can take arbitrary commands to execute before or after its rotations. Just as a basic example from the default logrotate configuration in CentOS, you can see this configuration:

Advertisement
free widgets for website
 /var/log/cron /var/log/maillog /var/log/messages /var/log/secure /var/log/spooler {     sharedscripts     postrotate         /bin/kill -HUP `cat /var/run/syslogd.pid 2> /dev/null` 2> /dev/null || true     endscript } 

A bit brittle, but we’ll forgive that and assume that this works as intended. This configuration says that after logrotate rotates any of the files listed, it should send SIGHUP to the pid contained in /var/run/syslogd.pid, which should be that of the running syslogd instance.

This is all well and good for something with a stable public API like syslog, but what about something internal where the implementation of SIGHUP is an internal implementation detail that could change at any time?

A History of Hangups

One of the problems here is that, except for signals which cannot be caught in user space and thus have only one meaning, like SIGKILL and SIGSTOP, the semantic meaning of signals is up to application developers and users to interpret and program. In some cases, the distinction is largely academic, like SIGTERM, which is pretty much universally understood to mean “terminate gracefully as soon as possible.” However, in the case of SIGHUP, the meaning is significantly less clear.

SIGHUP was invented for serial lines and was originally used to indicate that the other end of the connection had dropped the line. Nowadays, we still carry our lineage with us of course, so SIGHUP is still sent for its modern equivalent: where a pseudo or virtual terminal is closed (hence tools like nohup, which mask it).

In the early days of Unix, there was a need to implement daemon reloading. This usually consists at least of configuration/log file reopening without restarting, and signals seemed like a dependency-free way to achieve that. Of course, there was no signal for such a thing, but as these daemons have no controlling terminal, there should be no reason to receive SIGHUP, so it seemed like a convenient signal to piggyback onto without any obvious side effects.

Advertisement
free widgets for website

There is a small hitch with this plan though. The default state for signals is not “ignored,” but signal-specific. So, for example, programs don’t have to configure SIGTERM manually to terminate their application. As long as they don’t set any other signal handler, the kernel just terminates their program for free, without any code needed in user space. Convenient!

What’s not so convenient though, is that SIGHUP also has the default behavior of terminating the program immediately. This works great for the original hangup case, where these applications likely aren’t needed anymore, but is not so great for this new meaning.

This would be fine of course, if we removed all the places which could potentially send SIGHUP to the program. The problem is that in any large, mature codebase, that is difficult. SIGHUP is not like a tightly controlled IPC call for which you can easily grep the codebase for. Signals can come from anywhere, at any time, and there are few checks on their operation (other than the most basic “are you this user or have CAP_KILL“). The bottom line is that it’s hard to determine where signals could come from, but with more explicit IPC, we would know that this signal doesn’t mean anything to us and should be ignored.

See also  Facebook missed weeks of warning signs over Capitol attack, documents suggest - The Guardian

From Hangup to Hazard

By now, I suppose you may have started to guess what happened. A LogDevice release started one fateful afternoon containing the aforementioned code change. At first, nothing had gone awry, but at midnight the next day, everything mysteriously started falling over. The reason is the following stanza in the machine’s logrotate configuration, which sends a now unhandled (and therefore fatal) SIGHUP to the logdevice daemon:

 /var/log/logdevice/audit.log {   daily   # [...]   postrotate     pkill -HUP logdeviced   endscript } 

Missing just one short stanza of a logrotate configuration is incredibly easy and common when removing a large feature. Unfortunately, it’s also hard to be certain that every last vestige of its existence was removed at once. Even in cases that are easier to validate than this, it’s common to mistakenly leave remnants when doing code cleanup. Still, usually, it’s without any destructive consequences, that is, the remaining detritus is just dead or no-op code.

Advertisement
free widgets for website

Conceptually, the incident itself and its resolution are simple: don’t send SIGHUP, and spread LogDevice actions out more over time (that is, don’t run this at midnight on the dot). However, it’s not just this one incident’s nuances that we should focus on here. This incident, more than anything, has to serve as a platform to discourage the use of signals in production for anything other than the most basic, essential cases.

The Dangers of Signals

What Signals are Good For

First, using signals as a mechanism to affect changes in the process state of the operating system is well founded. This includes signals like SIGKILL, which are impossible to install a signal handler for and does exactly what you would expect, and the kernel-default behavior of SIGABRT, SIGTERM, SIGINT, SIGSEGV, and SIGQUIT and the like, which are generally well understood by users and programmers.

What these signals all have in common is that once you’ve received them, they’re all progressing towards a terminal end state within the kernel itself. That is, no more user space instructions will be executed once you get a SIGKILL or SIGTERM with no user space signal handler.

A terminal end state is important because it usually means you’re working towards decreasing the complexity of the stack and code currently being executed. Other desired states often result in the complexity actually becoming higher and harder to reason about as concurrency and code flow become more muddled.

Dangerous Default Behavior

You may notice that we didn’t mention some other signals that also terminate by default. Here’s a list of all of the standard signals that terminate by default (excluding core dump signals like SIGABRT or SIGSEGV, since they’re all sensible):

Advertisement
free widgets for website
  • SIGALRM
  • SIGEMT
  • SIGHUP
  • SIGINT
  • SIGIO
  • SIGKILL
  • SIGLOST
  • SIGPIPE
  • SIGPOLL
  • SIGPROF
  • SIGPWR
  • SIGSTKFLT
  • SIGTERM
  • SIGUSR1
  • SIGUSR2
  • SIGVTALRM

At first glance, these may seem reasonable, but here are a few outliers:

  • SIGHUP: If this was used only as it was originally intended, defaulting to terminate would be sensible. With the current mixed usage meaning “reopen files,” this is dangerous.
  • SIGPOLL and SIGPROF: These are in the bucket of “these should be handled internally by some standard function rather than your program.” However, while probably harmless, the default behavior to terminate still seems nonideal.
  • SIGUSR1 and SIGUSR2: These are “user-defined signals” that you can ostensibly use however you like. But because these are terminal by default, if you implement USR1 for some specific need and later don’t need that, you can’t just safely remove the code. You have to consciously think to explicitly ignore the signal. That’s really not going to be obvious even to every experienced programmer.

So that’s almost one-third of terminal signals, which are at best questionable and, at worst, actively dangerous as a program’s needs change. Worse still, even the supposedly “user-defined” signals are a disaster waiting to happen when someone forgets to explicitly SIG_IGN it. Even an innocuous SIGUSR1 or SIGPOLL may cause incidents.

This is not simply a question of familiarity. No matter how well you know how signals work, it’s still extremely hard to write signal-correct code the first time around because, despite their appearance, signals are far more complex than they seem.

Code flow, Concurrency, and the Myth of SA_RESTART

Programmers generally do not spend their entire day thinking about the inner workings of signals. This means that when it comes to actually implementing signal handling, they often subtly do the wrong thing.

I’m not even talking about the “trivial” cases, like safety in a signal handling function, which is mostly solved by only bumping a sig_atomic_t, or using C++’s atomic signal fence stuff. No, that’s mostly easily searchable and memorable as a pitfall by anyone after their first time through signal hell. What’s a lot harder is reasoning about the code flow of the nominal portions of a complex program when it receives a signal. Doing so requires either constantly and explicitly thinking about signals at every part of the application life cycle (hey, what about EINTR, is SA_RESTART enough here? What flow should we go into if this terminates prematurely? I now have a concurrent program, what are the implications of that?), or setting up a sigprocmask or pthread_setmask for some part of your application life cycle and praying that the code flow never changes (which is certainly not a good guess in an atmosphere of fast-paced development). signalfd or running sigwaitinfo in a dedicated thread can help somewhat here, but both of these have enough edge cases and usability concerns to make them hard to recommend.

We like to believe that most experienced programmers know by now that even a facetious example of correctly writing thread-safe code is very hard. Well, if you thought correctly writing thread-safe code was hard, signals are significantly harder. Signal handlers must only rely on strictly lock-free code with atomic data structures, respectively, because the main flow of execution is suspended and we don’t know what locks it’s holding, and because the main flow of execution could be performing non-atomic operations. They must also be fully reentrant, that is, they must be able to nest within themselves since signal handlers can overlap if a signal is sent multiple times (or even with one signal, with SA_NODEFER). That’s one of the reasons why you can’t use functions like printf or malloc in a signal handler because they rely on global mutexes for synchronization. If you were holding that lock when the signal was received and then called a function requiring that lock again, your application would end up deadlocked. This is really, really hard to reason about. That’s why many people simply write something like the following as their signal handling:

 static volatile sig_atomic_t received_sighup;   static void sighup(int sig __attribute__((unused))) { received_sighup = 1; }  static int configure_signal_handlers(void) {   return sigaction(     SIGHUP,     &(const struct sigaction){.sa_handler = sighup, .sa_flags = SA_RESTART},     NULL); }  int main(int argc, char *argv[]) {   if (configure_signal_handlers()) {        /* failed to set handlers */   }    /* usual program flow */    if (received_sighup) {     /* reload */     received_sighup = 0;   }    /* usual program flow */ }  

The problem is that, while this, signalfd, or other attempts at async signal handling might look fairly simple and robust, it ignores the fact that the point of interruption is just as important as the actions performed after receiving the signal. For example, suppose your user space code is doing I/O or changing the metadata of objects that come from the kernel (like inodes or FDs). In this case, you’re probably actually in a kernel space stack at the time of interruption. For example, here’s how a thread might look when it’s trying to close a file descriptor:

Advertisement
free widgets for website
# cat /proc/2965230/stack  [<0>] schedule+0x43/0xd0  [<0>] io_schedule+0x12/0x40  [<0>] wait_on_page_bit+0x139/0x230  [<0>] filemap_write_and_wait+0x5a/0x90  [<0>] filp_close+0x32/0x70  [<0>] __x64_sys_close+0x1e/0x50  [<0>] do_syscall_64+0x4e/0x140  [<0>] entry_SYSCALL_64_after_hwframe+0x44/0xa9

Here, __x64_sys_close is the x86_64 variant of the close system call, which closes a file descriptor. At this point in its execution, we’re waiting for the backing storage to be updated (that’s this wait_on_page_bit). Since I/O work is usually several orders of magnitude slower than other operations, schedule here is a way of voluntarily hinting to the kernel’s CPU scheduler that we are about to perform a high-latency operation (like disk or network I/O) and that it should consider finding another process to schedule instead of the current process for now. This is good, as it allows us to signal to the kernel that it is a good idea to go ahead and pick a process that will actually make use of the CPU instead of wasting time on one which can’t continue until it’s finished waiting for a response from something that may take a while.

Imagine that we send a signal to the process we were running. The signal that we have sent has a user space handler in the receiving thread, so we’ll resume in user space. One of the many ways this race can end up is that the kernel will try to come out of schedule, further unwind the stack and eventually return an errno of ESYSRESTART or EINTR to user space to indicate that we were interrupted. But how far did we get in closing it? What’s the state of the file descriptor now?

Now that we’ve returned to user space, we’ll run the signal handler. When the signal handler exits, we’ll propagate the error to the user space libc’s close wrapper, and then to the application, which, in theory, can do something about the situation encountered. We say “in theory” because it’s really hard to know what to do about many of these situations with signals, and many services in production do not handle the edge cases here very well. That might be fine in some applications where data integrity isn’t that important. However, in production applications that do care about data consistency and integrity, this presents a significant problem: the kernel doesn’t expose any granular way to understand how far it got, what it achieved and didn’t and what we should actually do about the situation. Even worse, if close returns with EINTR, the state of the file descriptor is now unspecified:

“If close() is interrupted by a signal [...] the state of [the file descriptor] is unspecified.”

Good luck trying to reason about how to handle that safely and securely in your application. In general, handling EINTR even for well-behaved syscalls is complicated. There are plenty of subtle issues forming a large part of the reason why SA_RESTART is not enough. Not all system calls are restartable, and expecting every single one of your application’s developers to understand and mitigate the deep nuances of getting a signal for every single syscall at every single call site is asking for outages. From man 7 signal:

Advertisement
free widgets for website

“The following interfaces are never restarted after being interrupted by a signal handler, regardless of the use of SA_RESTART; they always fail with the error EINTR [...]”

Likewise, using a sigprocmask and expecting code flow to remain static is asking for trouble as developers do not typically spend their lives thinking about the bounds of signal handling or how to produce or preserve signal-correct code. The same goes for handling signals in a dedicated thread with sigwaitinfo, which can easily end up with GDB and similar tools being unable to debug the process. Subtly wrong code flows or error handling can result in bugs, crashes, difficult to debug corruptions, deadlocks and many more issues that will send you running straight into the warm embrace of your preferred incident management tool.

High Complexity in Multithreaded Environments

If you thought all this talk of concurrency, reentrancy and atomicity was bad enough, throwing multithreading into the mix makes things even more complicated. This is especially important when considering the fact that many complex applications run separate threads implicitly, for example, as part of jemalloc, GLib, or similar. Some of these libraries even install signal handlers themselves, opening a whole other can of worms.

Overall, man 7 signal has this to say on the matter:

“A signal may be generated (and thus pending) for a process as a whole (e.g., when sent using kill(2)) or for a specific thread [...] If more than one of the threads has the signal unblocked, then the kernel chooses an arbitrary thread to which to deliver the signal.”

Advertisement
free widgets for website

More succinctly, “for most signals, the kernel sends the signal to any thread that doesn’t have that signal blocked with sigprocmask“. SIGSEGV, SIGILL and the like resemble traps, and have the signal explicitly directed at the offending thread. However, despite what one might think, most signals cannot be explicitly sent to a single thread in a thread group, even with tgkill or pthread_kill.

This means that you can’t trivially change overall signal handling characteristics as soon as you have a set of threads. If a service needs to do periodic signal blocking with sigprocmask in the main thread, you need to somehow communicate to other threads externally about how they should handle that. Otherwise, the signal may be swallowed by another thread, never to be seen again. Of course, you can block signals in child threads to avoid this, but if they need to do their own signal handling, even for primitive things like waitpid, it will end up making things complex.

Just as with everything else here, these aren’t technically insurmountable problems. However, one would be negligent in ignoring the fact that the complexity of synchronization required to make this work correctly is burdensome and lays the groundwork for bugs, confusion and worse.

Lack of Definition and Communication of Success or Failure

Signals are propagated asynchronously in the kernel. The kill syscall returns as soon as the pending signal is recorded for the process or thread’s task_struct in question. Thus, there’s no guarantee of timely delivery, even if the signal isn’t blocked.

Even if there is timely delivery of the signal, there’s no way to communicate back to the signal issuer what the status of their request for action is. As such, any meaningful action should not be delivered by signals, since they only implement fire-and-forget with no real mechanism to report the success or failure of delivery and subsequent actions. As we’ve seen above, even seemingly innocuous signals can be dangerous when they are not configured in user space.

Advertisement
free widgets for website

Anyone using Linux for long enough has undoubtedly run into a case where they want to kill some process but find that the process is unresponsive even to supposedly always fatal signals like SIGKILL. The problem is that misleadingly, kill(1)’s purpose isn’t to kill processes, but just to queue a request to the kernel (with no indication about when it will be serviced) that someone has requested some action to be taken.

The kill syscall’s job is to mark the signal as pending in the kernel’s task metadata, which it does successfully even when a SIGKILL task doesn’t die. In the case of SIGKILL in particular, the kernel guarantees that no more user mode instructions will be executed, but we may still have to execute instructions in kernel mode to complete actions that otherwise may result in data corruption or to release resources. For this reason, we still succeed even if the state is D (uninterruptible sleep). Kill itself doesn’t fail unless you provided an invalid signal, you don’t have permission to send that signal or the pid that you requested to send a signal to does not exist and is thus not useful to reliably propagate non-terminal states to applications.

In Conclusion

  • Signals are fine for terminal state handled purely in-kernel with no user space handler. For signals that you actually would like to immediately kill your program, leave those signals alone for the kernel to handle. This also means that the kernel may be able to exit early from its work, freeing up your program resources more quickly, whereas a user space IPC request would have to wait for the user space portion to start executing again.
  • A way to avoid getting into trouble handling signals is to not handle them at all. However, for applications handling state processing that must do something about cases like SIGTERM, ideally use a high-level API like folly::AsyncSignalHandler where a number of the warts have already been made more intuitive.

  • Avoid communicating application requests with signals. Use self-managed notifications (like inotify) or user space RPC with a dedicated part of the application life cycle to handle it instead of relying on interrupting the application.
  • Where possible, limit the scope of signals to a subsection of your program or threads with sigprocmask, reducing the amount of code that needs to be regularly scrutinized for signal-correctness. Bear in mind that if code flows or threading strategies change, the mask may not have the effect you intended.
  • At daemon start, mask terminal signals that are not uniformly understood and could be repurposed at some point in your program to avoid falling back to kernel default behavior. My suggestion is the following:
 signal(SIGHUP, SIG_IGN); signal(SIGQUIT, SIG_IGN); signal(SIGUSR1, SIG_IGN); signal(SIGUSR2, SIG_IGN); 

Signal behavior is extremely complicated to reason about even in well-authored programs, and its use presents an unnecessary risk in applications where other alternatives are available. In general, do not use signals for communicating with the user space portion of your program. Instead, either have the program transparently handle events itself (for example, with inotify), or use user space communication that can report back errors to the issuer and is enumerable and demonstrable at compile time, like Thrift, gRPC or similar.

I hope this article has shown you that signals, while they may ostensibly appear simple, are in reality anything but. The aesthetics of simplicity that promote their use as an API for user space software belie a series of implicit design decisions that do not fit most production use cases in the modern era.

Let’s be clear: there are valid use cases for signals. Signals are fine for basic communication with the kernel about a desired process state when there’s no user space component, for example, that a process should be killed. However, it is difficult to write signal-correct code the first time around when signals are expected to be trapped in user space.

Signals may seem attractive due to their standardization, wide availability and lack of dependencies, but they come with a significant number of pitfalls that will only increase concern as your project grows. Hopefully, this article has provided you with some mitigations and alternative strategies that will allow you to still achieve your goals, but in a safer, less subtly complex and more intuitive way.

Advertisement
free widgets for website

To learn more about Meta Open Source, visit our open source site, subscribe to our YouTube channel, or follow us on Twitter, Facebook and LinkedIn.

First seen at developers.facebook.com

Continue Reading

FACEBOOK

Meet the Developers – Linux Kernel Team (David Vernet)

Published

on

By

meet-the-developers-–-linux-kernel-team-(david-vernet)

Credit: Larry Ewing (lewing@isc.tamu.edu) and The GIMP for the original design of Tux the penguin.

Intro

For today’s interview, we have David Vernet, a core systems engineer on the Kernel team at Meta. He works on the BPF (Berkeley Packet Filter) and the Linux kernel scheduler. This series highlights Meta Software Engineers who contribute to the Linux kernel. The Meta Linux Kernel team works with the broader Linux community to add new features to the kernel and makes sure that the kernel works well in Meta production data centers. Engineers on the team work with peers in the industry to make the kernel better for Meta’s workloads and to make Linux better for everyone.

Tell us about yourself.

I’m a systems engineer who’s spent a good chunk of his career in the kernel space, and some time in the user-space as well working on a microkernel. Right now, I’m focusing most of my time on BPF and the Linux kernel scheduler.

I started my career as a web developer after getting a degree in math. After going to grad school, I realized that I was happiest when hacking on low-level systems and figuring out how computers work.

As a kernel developer at Meta, what does your typical day look like?

I’m not a maintainer of any subsystems in the kernel, so my typical day is filled with almost exclusively coding and engineering. That being said, participating in the upstream Linux kernel community is one of the coolest parts of being on the kernel team, so I still spend some time reading over upstream discussions. A typical day goes something like this:

Advertisement
free widgets for website
  1. Read over some of the discussions taking place on various upstream lists, such as BPF and mm. I usually spend about 30-60 minutes or so per day on this, though it depends on the day.

  2. Hack on the project that I’m working on. Lately, that’s adding a user-space ringbuffer map type to BPF.

  3. Work on drafting an article for lwn.net.

What have you been excited about or incredibly proud of lately?

I recently submitted a patch-set to enable a new map type in BPF. This allows user-space to publish messages to BPF programs in the kernel over the ringbuffer. This map type is exciting because it sets the stage to enable frameworks for user-space to drive logic in BPF programs in a performant way.

Is there something especially exciting about being a kernel developer at a company like Meta?

The Meta kernel team has a strong upstream-first culture. Bug fixes that we find in our Meta kernel, and features that we’d like to add, are almost always first submitted to the upstream kernel, and then they are backported to our internal kernel.

Do you have a favorite part of the kernel dev life cycle?

I enjoy architecting and designing APIs. Kernel code can never crash and needs to be able to run forever. I find it gratifying to architect systems in the kernel that make it easy to reason about correctness and robustness and provide intuitive APIs that make it easy for other parts of the kernel to use your code.

I also enjoy iterating with the upstream community. It’s great that your patches have a whole community of people looking at them to help you find bugs in your code and suggest improvements that you may never have considered on your own. A lot of people find this process to be cumbersome, but I find that it’s a small price to pay for what you get out of it.

Tell us a bit about the topic you presented at the Linux Plumbers Conference this year.

We presented the live patch feature in the Linux kernel, describing how we have utilized it at Meta and how our hyper-scale has shown some unique challenges with the feature.

Advertisement
free widgets for website

What are some of the misconceptions about kernel or open source software development that you have encountered in your career?

The biggest misconception is that it’s an exclusive, invite-only club to contribute to the Linux kernel. You certainly must understand operating systems to be an effective contributor and be ready to receive constructive criticism when there is scope for improvement in your code. Still, the community always welcomes people who come in with an open mind and want to contribute.

What resources are helpful in getting started in kernel development?

There is a lot of information out there that people have written on how to get integrated into the Linux kernel community. I wrote a blog post on how to get plugged into Linux kernel upstream mailing list discussions, and another on how to submit your first patch. There is also a video on writing and submitting your first Linux kernel patch from Greg Kroah-Hartman.

In terms of resources to learn about the kernel itself, there are many resources and books, such as:

Where can people find you and follow your work?

I have a blog where I talk about my experiences as a systems engineer: https://www.bytelab.codes/. I publish articles that range from topics that are totally newcomer friendly to more advanced topics that discuss kernel code in more detail. Feel free to check it out and let me know if there’s anything you’d like me to discuss.

To learn more about Meta Open Source, visit our open source site, subscribe to our YouTube channel, or follow us on Twitter, Facebook and LinkedIn.

First seen at developers.facebook.com

Advertisement
free widgets for website
See also  Facebook Teases Technology That Allows Your Brain To Control A Computer
Continue Reading

Trending