Language Selection

English French German Italian Portuguese Spanish

Kernel Planet

Syndicate content
Kernel Planet - http://planet.kernel.org
Updated: 4 hours 9 min ago

Kees Cook: security things in Linux v5.7

Monday 21st of September 2020 11:32:16 PM

Previously: v5.6

Linux v5.7 was released at the end of May. Here’s my summary of various security things that caught my attention:

arm64 kernel pointer authentication
While the ARMv8.3 CPU “Pointer Authentication” (PAC) feature landed for userspace already, Kristina Martsenko has now landed PAC support in kernel mode. The current implementation uses PACIASP which protects the saved stack pointer, similar to the existing CONFIG_STACKPROTECTOR feature, only faster. This also paves the way to sign and check pointers stored in the heap, as a way to defeat function pointer overwrites in those memory regions too. Since the behavior is different from the traditional stack protector, Amit Daniel Kachhap added an LKDTM test for PAC as well.

BPF LSM
The kernel’s Linux Security Module (LSM) API provide a way to write security modules that have traditionally implemented various Mandatory Access Control (MAC) systems like SELinux, AppArmor, etc. The LSM hooks are numerous and no one LSM uses them all, as some hooks are much more specialized (like those used by IMA, Yama, LoadPin, etc). There was not, however, any way to externally attach to these hooks (not even through a regular loadable kernel module) nor build fully dynamic security policy, until KP Singh landed the API for building LSM policy using BPF. With this, it is possible (for a privileged process) to write kernel LSM hooks in BPF, allowing for totally custom security policy (and reporting).

execve() deadlock refactoring
There have been a number of long-standing races in the kernel’s process launching code where ptrace could deadlock. Fixing these has been attempted several times over the last many years, but Eric W. Biederman and Ernd Edlinger decided to dive in, and successfully landed the a series of refactorings, splitting up the problematic locking and refactoring their uses to remove the deadlocks. While he was at it, Eric also extended the exec_id counter to 64 bits to avoid the possibility of the counter wrapping and allowing an attacker to send arbitrary signals to processes they normally shouldn’t be able to.

slub freelist obfuscation improvements
After Silvio Cesare observed some weaknesses in the implementation of CONFIG_SLAB_FREELIST_HARDENED‘s freelist pointer content obfuscation, I improved their bit diffusion, which makes attacks require significantly more memory content exposures to defeat the obfuscation. As part of the conversation, Vitaly Nikolenko pointed out that the freelist pointer’s location made it relatively easy to target too (for either disclosures or overwrites), so I moved it away from the edge of the slab, making it harder to reach through small-sized overflows (which usually target the freelist pointer). As it turns out, there were a few assumptions in the kernel about the location of the freelist pointer, which had to also get cleaned up.

RISCV page table dumping
Following v5.6’s generic page table dumping work, Zong Li landed the RISCV page dumping code. This means it’s much easier to examine the kernel’s page table layout when running a debug kernel (built with PTDUMP_DEBUGFS), visible in /sys/kernel/debug/kernel_page_tables.

array index bounds checking
This is a pretty large area of work that touches a lot of overlapping elements (and history) in the Linux kernel. The short version is: C is bad at noticing when it uses an array index beyond the bounds of the declared array, and we need to fix that. For example, don’t do this:

int foo[5]; ... foo[8] = bar;

The long version gets complicated by the evolution of “flexible array” structure members, so we’ll pause for a moment and skim the surface of this topic. While things like CONFIG_FORTIFY_SOURCE try to catch these kinds of cases in the memcpy() and strcpy() family of functions, it doesn’t catch it in open-coded array indexing, as seen in the code above. GCC has a warning (-Warray-bounds) for these cases, but it was disabled by Linus because of all the false positives seen due to “fake” flexible array members. Before flexible arrays were standardized, GNU C supported “zero sized” array members. And before that, C code would use a 1-element array. These were all designed so that some structure could be the “header” in front of some data blob that could be addressable through the last structure member:

/* 1-element array */ struct foo { ... char contents[1]; }; /* GNU C extension: 0-element array */ struct foo { ... char contents[0]; }; /* C standard: flexible array */ struct foo { ... char contents[]; }; instance = kmalloc(sizeof(struct foo) + content_size);

Converting all the zero- and one-element array members to flexible arrays is one of Gustavo A. R. Silva’s goals, and hundreds of these changes started landing. Once fixed, -Warray-bounds can be re-enabled. Much more detail can be found in the kernel’s deprecation docs.

However, that will only catch the “visible at compile time” cases. For runtime checking, the Undefined Behavior Sanitizer has an option for adding runtime array bounds checking for catching things like this where the compiler cannot perform a static analysis of the index values:

int foo[5]; ... for (i = 0; i < some_argument; i++) { ... foo[i] = bar; ... }

It was, however, not separate (via kernel Kconfig) until Elena Petrova and I split it out into CONFIG_UBSAN_BOUNDS, which is fast enough for production kernel use. With this enabled, it's now possible to instrument the kernel to catch these conditions, which seem to come up with some regularity in Wi-Fi and Bluetooth drivers for some reason. Since UBSAN (and the other Sanitizers) only WARN() by default, system owners need to set panic_on_warn=1 too if they want to defend against attacks targeting these kinds of flaws. Because of this, and to avoid bloating the kernel image with all the warning messages, I introduced CONFIG_UBSAN_TRAP which effectively turns these conditions into a BUG() without needing additional sysctl settings.

Fixing "additive" snprintf() usage
A common idiom in C for building up strings is to use sprintf()'s return value to increment a pointer into a string, and build a string with more sprintf() calls:

/* safe if strlen(foo) + 1 < sizeof(string) */ wrote = sprintf(string, "Foo: %s\n", foo); /* overflows if strlen(foo) + strlen(bar) > sizeof(string) */ wrote += sprintf(string + wrote, "Bar: %s\n", bar); /* writing way beyond the end of "string" now ... */ wrote += sprintf(string + wrote, "Baz: %s\n", baz);

The risk is that if these calls eventually walk off the end of the string buffer, it will start writing into other memory and create some bad situations. Switching these to snprintf() does not, however, make anything safer, since snprintf() returns how much it would have written:

/* safe, assuming available <= sizeof(string), and for this example * assume strlen(foo) < sizeof(string) */ wrote = snprintf(string, available, "Foo: %s\n", foo); /* if (strlen(bar) > available - wrote), this is still safe since the * write into "string" will be truncated, but now "wrote" has been * incremented by how much snprintf() *would* have written, so "wrote" * is now larger than "available". */ wrote += snprintf(string + wrote, available - wrote, "Bar: %s\n", bar); /* string + wrote is beyond the end of string, and availabe - wrote wraps * around to a giant positive value, making the write effectively * unbounded. */ wrote += snprintf(string + wrote, available - wrote, "Baz: %s\n", baz);

So while the first overflowing call would be safe, the next one would be targeting beyond the end of the array and the size calculation will have wrapped around to a giant limit. Replacing this idiom with scnprintf() solves the issue because it only reports what was actually written. To this end, Takashi Iwai has been landing a bunch scnprintf() fixes.

That's it for now! Let me know if there is anything else you think I should mention here. Next up: Linux v5.8.

© 2020, Kees Cook. This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 License.

James Bottomley: Creating a Home IPv6 Network

Thursday 17th of September 2020 10:23:20 PM

One of the recent experiences of Linux Plumbers Conference convinced me that if you want to be part of a true open source WebRTC based peer to peer audio/video interaction, you need an internet address that’s not behind a NAT. In reality, the protocol still works as long as you can contact a stun server to tell you what your external address is and possibly a turn server to proxy the packets if both endpoints are NATed but all this seeking external servers takes time as those of you who complained about the echo test found. The solution to all this is to connect over IPv6 which has an address space large enough to support every device on the planet having its own address. All modern Linux distributions support IPv6 out of the box so the chances are you’ve actually accidentally used it without ever noticing, which is one of the beauties of IPv6 autoconfiguration (it’s supposed to just work).

However, I recently moved, and so lost my fibre internet connection to cable but cable that did come with an IPv6 address, so this is my story of getting it all to work. If you don’t really care about the protocol basics, you can skip down to the how. This guide is also focussed on a “dual stack” configuration (one that has both IPv6 and IPv4 addresses). Pure IPv6 configurations are possible, but because some parts of the internet are still IPv4 only, they’re not complete unless you set up an IPv4 encapsulating bridge.

The Basics of IPv6

IPv6 has been a mature protocol for a long time now, so I erroneously assumed there’d be a load of good HOWTOs about it. However, after reading 20 different descriptions of how the IPv6 128 bit address space works and not much else, I gave up in despair and read the RFCs instead. I’ll assume you’ve read at least one of these HOWTOS, so I don’t have to go into IPv6 address prefixes, suffixes, interface IDs or subnets so I’ll begin where most of the HOWTOs end.

How does IPv6 Just Work?

In IPv4 there’s a protocol called dynamic host configuration protocol (DHCP) so as long as you can find a DHCP server you can get all the information you need to connect (local address, router, DNS server, time server, etc). However, this service has to be set up by someone and IPv6 is designed to configure a network without it.

The first assumption IPv6 StateLess Address AutoConfiguration (SLAAC) makes is that it’s on a /64 subnet (So every subnet in IPv6 contains 1010 times as many addresses as the entire IPv4 internet). This means that, since most real subnets contain <100 systems, they can simply choose a random address and be very unlikely to clash with the existing systems. In fact, there are three current ways of choosing an address in the /64:

  1. EUI-64 (RFC 4291) based on the MAC address which is basically the MAC with one bit flipped and ff:fe placed in the middle.
  2. Stable Private (RFC 7217) which generate from a hash based on a static key, interface, prefix and a counter (the counter is incremented if there is a clash). These are preferred to the EUI-64 ones which give away any configuration associated with the MAC address (such as what type of network card you have)
  3. Privacy Extension Addresses (RFC 4941) which are very similar to stable private addresses except they change over time using the IPv6 address deprecation mechanism and are for client systems who want to preserve anonymity.

The next problem in Linux is who configures the interface? The Kernel IPv6 stack is actually designed to do it, and will unless told not to, but most of the modern network controllers (like NetworkManager) are control freaks and turn off the kernel’s auto configuration so they can do it themselves. They also default to stable private addressing using a static secret maintained in the filesystem (/var/lib/NetworkManager/secret_key).

The next thing to understand about IPv6 addresses is that they are divided into scopes, the most important being link local (unrouteable) addresses which conventionally always have the prefix fe80::/64. The link local address is configured first using one of the above methods and then used to probe the network.

Multicast and Neighbour Discovery

Unlike IPv4, IPv6 has no broadcast capability so all discovery is done by multicast. Nodes coming up on the network subscribe to particular multicast addresses, via special packets intercepted by the switch, and won’t receive any multicast to which they’re not subscribed. Conventionally, all link local multicast addresses have the prefix ff02::/64 (for other types of multicast address see RFC 4291). All nodes subscribe to the “all nodes” multicast address ff02::1 and also must subscribe to their own solicited node multicast address at ff02::1:ffXX:XXXX where the last 24 bits correspond to the lowest 24 bits of the node’s IPv6 address. This latter is to avoid the disruption that used to occur in IPv4 from ARP broadcasts because now you can target a specific subset of nodes for address resolution.

The IPV6 address resolution protocol is called Neighbour Solicitation (NS), described in RFC 4861 and it’s use with SLAAC described in RFC 4862, and is done by sending a multicast to the neighbor solicitation address of the node you want to discover containing the full IPv6 address you want to know, a node with the matching address replies with its link layer (MAC) address in a Neighbour Advertisement (NA) packet.

Once a node has chosen its link local address, it first sends out a NS packet to its chosen address to see if anyone replies and if no-one does it assumes it is OK to keep it otherwise it follows the collision avoidance protocol associated with its particular form of address. Once it has found a unique address, the node configures this link local address and looks for a router. Note that if an IPv6 network isn’t present, discovery stops here, which is why most network interfaces always show a link local IPv6 address.

Router Discovery

Once the node has its own unique link local address, it uses it to send out Router Solicitation (RS) packets to the “all routers” multicast address ff02::2. Every router on the network responds with a Router Advertisement (RA) packet which describes (among other things) the the router lifetime, the network MTU, a set of one or more prefixes the router is responsible for, the router’s link address and a set of option flags including the M (Managed) and O (Other Configuration) flag and possibly a set of DNS servers.

Each advertised prefix contains the prefix and prefix length, a set of flags including the A (autonomous configuration) and L (link local) and a set of lifetimes. The Link Local prefixes tell you what global prefixes the local network users (there may be more than one) and whether you are allowed to do SLAAC on the global prefix (if the A flag is clear, you must ask the router for an address using DHCPv6). If the router has a non zero lifetime, you may assume it is a default router for the subnet.

Now that the node has discovered one or more routers it may configure its own global address (note that every IPv6 routeable node has at least two addresses: a link local and a global). How it does this depends on the router and prefix flags

Global Address Configuration

The first thing a node needs to know is whether to use SLAAC for the global address or DHCPv6. This is entirely determined by the A flag of any link local prefix in the RA packet. If A is set, then the node may use SLAAC and if A is clear then the node must use DHCPv6 to obtain an address. If A is set and also the M (Managed) flag then the node may use either SLAAC or DHCPv6 (or both) to obtain an address and if the M flag is clear, but the O (Other Config) flag is present then the node must use SLAAC but may use DHCPv6 to obtain other information about the network (usually DNS).

Once the node has a global address in now needs a default route. It forms the default route list from the RA packets that have a non-zero router Lifetime. All of these are configured as default routes to their link local address with the RA specified hop count. Finally, the node may add specific prefix routes from RA packets with zero router LifeTimes but non link local prefixes.

DHCPv6 is a fairly complex configuration protocol (see RFC 8415) but it cannot specify either prefix length (meaning all obtained addresses are configured as /128) or routes (these must be obtained from RA packets). This leads to a subtlety of outbound address selection in that the most specific is always preferred, so if you configure both by SLAAC and DHCPv6, the SLAAC address will be added as /64 and the DHCPv6 address as /128 meaning your outbound IP address will always be the DHCPv6 one (although if an external entity knows your SLAAC address, they will still be able to reach you on it).

The How: Configuring your own Home Router

One of the things you’d think from the above is that IPv6 always auto configures and, while it is true that if you simply plug your laptop into the ethernet port of a cable modem it will just automatically configure, most people have a more complex home setup involving a router, which needs some special coaxing before it will work. That means you need to obtain additional features from your ISP using special DHCPv6 requests.

This section is written from my own point of view: I have a rather complex IPv4 network which has a completely open but bandwidth limited (to untrusted clients) wifi network, and several protected internal networks (one for my lab, one for my phones and one for the household video cameras), so I need at least 4 subnets to give every device in my home an IPv6 address. I also use OpenWRT as my router distribution, so all the IPv6 configuration information is highly specific to it (although it should be noted that things like NetworkManager can also do all of this if you’re prepared to dig in the documentation).

Prefix Delegation

Since DHCPv6 only hands out a /128 address, this isn’t sufficient because it’s the IP address of the router itself. In order to become a router, you must request delegation of part of the IPv6 address space via the Identity Association for Prefix Delection (IA_PD) option of DHCPv6. Once this is done the router IP address will be assumed by the ISP to be the route for all of the delegated prefixes. The subtlety here is that if you want more than one subnet, you have to ask for it specifically (the client must specify the exact prefix length it’s looking for) and since it’s a prefix length, and your default subnet should be /64, if you request a prefix length of 64 you only have one subnet. If you request 63 you have 2 and so on. The problem is how do you know how many subnets the ISP is willing to give you? Unfortunately there’s no way of finding this (I had to do an internet search to discover my ISP, Comcast, was willing to delegate a prefix length of 60, meaning 16 subnets). If searching doesn’t tell you how much your ISP is willing to delegate, you could try starting at 48 and working your way to 64 in increments of 1 to see what the largest delegation you can get away with (There have been reports of ISPs locking you at your first delegated prefix length, so don’t start at 64). The final subtlety is that the prefix you’re delegated may not be the same prefix as the address your router obtained (my current comcast configuration has my router at 2001:558:600a:… but my delegated prefix is 2601:600:8280:66d0:/60). Note you can run odhcp6c manually with the -P option if you have to probe your ISP to find out what size of prefix you can get.

Configuring the Router for Prefix Delegation

In OpenWRT terms, the router WAN DHCP(v6) configuration is controlled by /etc/default/network. You’ll already have a WAN interface (likely called ‘wan’) for DHCPv4, so you simply add an additional ‘wan6’ interface to get an additional IPv6 and become dual stack. In my configuration this looks like

config interface 'wan6' option ifname '@wan' option proto 'dhcpv6' option reqprefix 60

The slight oddity is the ifname: @wan simply tells the config to use the same ifname as the ‘wan’ interface. Naming it this way is essential if your wan is a bridge, but it’s good practice anyway. The other option ‘reqprefix’ tells DHCPv6 to request a /60 prefix delegation.

Handing Out Delegated Prefixes

This turns out to be remarkably simple. Firstly you have to assign a delegated prefix to each of your other interfaces on the router, but you can do this without adding a new OpenWRT interface for each of them. My internal IPv4 network has all static addresses, so you add three directives to each of the interfaces:

config interface 'lan' ... interface designation (bridge for me) option proto 'static' ... ipv4 addresses option ip6assign '64' option ip6hint '1' option ip6ifaceid '::ff'

ip6assign ‘N’ means you are a /N network (so this is always /64 for me) and ip6hint ‘N’ means use N as your subnet id and ip6ifaceid ‘S’ means use S as the IPv6 suffix (This defaults to ::1 so if you’re OK with that, omit this directive). So given I have a 2601:600:8280:66d0::/60 prefix, the global address of this interface will be 2601:600:8280:66d1::ff. Now the acid test, if you got this right, this global address should be pingable from anywhere on the IPv6 internet (if it isn’t, it’s likely a firewall issue, see below).

Advertising as a Router

Simply getting delegated a delegated prefix on a local router interface is insufficient . Now you need to get your router to respond to Router Solicitations on ff02::2 and optionally do DHCPv6. Unfortunately, OpenWRT has two mechanisms for doing this, usually both installed: odhcpd and dnsmasq. What I found was that none of my directives in /etc/config/dhcp would take effect until I disabled odhcpd completely

/etc/init.d/odhcpd stop /etc/init.d/odhcpd disable

and since I use dnsmasq extensively elsewhere (split DNS for internal/external networks), that suited me fine. I’ll describe firstly what options you need in dnsmasq and secondly how you can achieve this using entries in the OpenWRT /etc/config/dhcp file (I find this useful because it’s always wise to check what OpenWRT has put in the /var/etc/dnsmasq.conf file).

The first dnsmasq option you need is ‘enable-ra’ which is a global parameter instructing dnsmasq to handle router advertisements. The next parameter you need is the per-interface ‘ra-param’ which specifies the global router advertisement parameters and must appear once for every interface you want to advertise on. Finally the ‘dhcp-range’ option allows more detailed configuration of the type of RA flags and optional DHCPv6.

SLAAC or DHCPv6 (or both)

In many ways this is a matter of personal choice. If you allow SLAAC, hosts which want to use privacy extension addresses (like Android phones) can do so, which is a good thing. If you also allow DHCPv6 address selection you will have a list of addresses assigned to hosts and dnsmasq will do DNS resolution for them (although it can do DNS for SLAAC addresses provided it gets told about them). A special tag ‘constructor’ exists for the ‘dhcp-range’ option which tells it to construct the supplied address (for either RA or DHCPv6) from the IPv6 global prefix of the specified interface, which is how you pass out our delegated prefix addresses. The modes for ‘dhcp-range’ are ‘ra-only’ to disallow DHCPv6 entirely, ‘slaac’ to allow DHCPv6 address selection and ‘ra-stateless’ to disallow DHCPv6 address selection but allow other DHCPv6 configuration information.

Based on trial and error (and finally examining the scripting in /etc/init.d/dnsmasq) the OpenWRT options required to achieve the above dnsmasq options are

config dhcp lan option interface lan option start 100 option limit 150 option leasetime 1h option dhcpv6 'server' option ra_management '1' option ra 'server'

with ‘ra_management’ as the key option with ‘0’ meaning SLAAC with DHCPv6 options, ‘1’ meaning SLAAC with full DHCPv6, ‘2’ meaning DHCPv6 only and ‘3’ meaning SLAAC only. Another OpenWRT oddity is that there doesn’t seem to be a way of setting the lease range: it always defaults to either static only or ::1000 to ::ffff.

Firewall Configuration

One of the things that trips people up is the fact that Linux has two completely separate firewalls: one for IPv4 and one for IPv6. If you’ve ever written any custom rules for them, the chances are you did it in the OpenWRT /etc/firewall.user file and you used the iptables command, which means you only added the rules to the IPv4 firewall. To add the same rule for IPv6 you need to duplicate it using the ip6tables command. Another significant problem, if you’re using a connection tracking for port knock detection like I am, is that Linux connection tracking has difficulty with IPv6 multicast, so packets that go out to a multicast but come back as unicast (as most of the discovery protocols do) get the wrong conntrack state. To fix this, I eventually had to have an INPUT rule just accepting all ICMPv6 and DHCPv6 (udp ports 546 [client] and 547 [server]). The other firewall considerations are that now everyone has their own IP address, there’s no need to NAT (OpenWRT can be persuaded to take care of this automatically, but if you’re duplicating the IPv4 rules manually, don’t duplicate the NAT rules). The final one is likely more applicable to me: my wifi interface is designed to be an extension of the local internet and all machines connecting to it are expected to be able to protect themselves since they’ll migrate to such hostile environments as airport wifi, thus I do complete exposure of wifi connected devices to the general internet for all ports, including port probes. For my internal devices, I have a RELATED,ESTABLISHED rule to make sure they’re not probed since they’re not designed to migrate off the internal network.

Now the problems with OpenWRT: since you want NAT on IPv4 but not on IPv6 you have to have two separate wan zones for them: if you try to combine them (as I first did), then OpenWRT will add an IPv6 –ctstate INVALID rule which will prevent Neighbour Discovery from working because of the conntrack issues with IPv6 multicast, so my wan zones are (well, this is a lie because my firewall is now hand crafted, but this is what I checked worked before I put the hand crafted firewall in place):

config zone option name 'wan' option network 'wan' option masq '1' ... config zone option name 'wan6' option network 'wan6' ...

And the routing rules for the lan zone (fully accessible) are

config forwarding option src 'lan' option dest 'wan' config forwarding option src 'lan' option dest 'wan6' config forwarding option src 'wan6' option dest 'lan' Putting it Together: Getting the Clients IPv6 Connected

Now that you have your router configured, everything should just work. If it did, your laptop wifi interface should now have a global IPv6 address

ip -6 address show dev wlan0

If that comes back empty, you need to enable IPv6 on your distribution. If it has only a link local (fe80:: prefix) address, IPv6 is enabled but your router isn’t advertising (suspect firewall issues with discovery packets or dnsmasq misconfiguration). If you see a global address, you’re done. Now you should be able to go to https://testv6.com and secure a 10/10 score.

The final piece of the puzzle is preferring your new IPv6 connection when DNS offers a choice of IPv4 or IPv6 addresses. All modern Linux clients should prefer IPv6 when available if connected to a dual stack network, so try … if you ping, say, www.google.com and see an IPv6 address you’re done. If not, you need to get into the murky world of IPv6 address labelling (RFC 6724) and gai.conf.

Conclusion

Adding IPv6 to and existing IPv4 setup is currently not a simple plug in and go operation. However, provided you understand a handful of differences between the two protocols, it’s not an insurmountable problem either. I have also glossed over many of the problems you might encounter with your ISP. Some people have reported that their ISPs only hand out one IPv6 address with no prefix delegation, in which case I think finding a new ISP would be wisest. Others report that the ISP will only delegate one /64 prefix so your choice here is either to only run one subnet (likely sufficient for a lot of home networks), or subnet at greater than /64 and forbid SLAAC, which is definitely not a recommended configuration. However, provided your ISP is reasonable, this blog post should at least help get you started.

Paul E. Mc Kenney: The Old Man and His Smartphone, 2020 “See You in September” Episode

Monday 7th of September 2020 04:02:39 AM
The continued COVID-19 situation continues to render my smartphone's location services less than useful, though a number of applications will still beg me to enable it, preferring to know my present location rather than consider my past habits. One in particular does have a “Don't ask me again” link, but it asks each time anyway. Given that I have only ever used one of that business's locations, you would think that it would not be all that hard to figure out which location I was going to be using next. But perhaps I am the only one who habitually disables location services.

Using the smartphone for breakfast-time Internet browsing has avoided almost all flat-battery incidents. One recent exception occurred while preparing for a hike. But I still have my old digital camera, so I plugged the smartphone into its charger and took my digital camera instead. I have previously commented on the excellent quality of my smartphone's cameras, but there is nothing quite like going back to the old digital camera (never mind my long-departed 35mm SLR) to drive that lesson firmly home.

I was recently asked to text a photo, and saw no obvious way to do this. There was some urgency, so I asked for an email address and emailed the photo instead. This did get the job done, but let's just say that it appears that asking for an email address is no longer a sign of youth, vigor, or with-it-ness. Thus chastened, I experimented in a calmer time, learning that the trick is to touch the greater-than icon to the left of the text-message-entry bar, which produces an option to select from your gallery and also to include a newly taken picture.

The appearance of Comet Neowise showcased my smartphone's ability to orient and to display the relevant star charts. Nevertheless, my wife expressed confidence in this approach only after seeing the large number of cars parked in the same area that my smartphone and I had selected. I hadn't intended to take a photo of the comet because the professionals a much better job, especially those who are willing to travel far away from city lights and low altitudes. But here was my smartphone and there was the comet, so why not? The resulting photo was quite unsatisfactory, with so much pixelated noise that the comet was just barely discernible.

It was some days later that I found the smartphone's night mode. This is quite impressive. In this mode, the smartphone can form low-light images almost as well as my eyes can, which is saying something. It is also extremely good with point sources of light.

One recent trend in clothing is pockets for smartphones. This trend prompted my stepfather to suggest that the smartphone is the pocket watch of the 21st century. This might well be, but I still wear a wristwatch.

My refusal to use my smartphone's location services does not mean that location services cannot get me in trouble. Far from it! One memorable incident took place on BPA Road in Forest Park. A group of hikers asked me to verify their smartphone's chosen route, which would have taken them past the end of Firelane 13 and eventually down a small cliff. I advised them to choose a different route.

But I had seen the little line that their smartphone had drawn, and a week or so later found myself unable to resist checking it out. Sure enough, when I peered through the shrubbery marking the end of Firelane 13, I saw an unassuming but very distinct trail. Of course I followed it. Isn't that what trails are for? Besides, maybe someone had found a way around the cliff I knew to be at the bottom of that route.

To make a long story short, no one had found a way around that cliff. Instead, the trail went straight down it. For all but about eight feet of the trail, it was possible to work my way down via convenient handholds in the form of ferns, bushes, and trees. My plan for that eight feet was to let gravity do the work, and to regain control through use of a sapling at the bottom of that stretch of the so-called trail. Fortunately for me, that sapling was looking out for this old man, but unfortunately this looking out took the form of ensuring that I had a subcutaneous hold on its bark. Thankfully, the remainder of the traverse down the cliff was reasonably uneventful.

Important safety tip: If you absolutely must use that trail, wear a pair of leather work gloves!

Paul E. Mc Kenney: Stupid RCU Tricks: Enlisting the Aid of a Debugger

Saturday 5th of September 2020 12:12:13 AM
Using Debuggers With rcutorture

So rcutorture found a bug, you have figured out how to reproduce it, git bisect was unhelpful (perhaps because the bug has been around forever), and the bug happens to be one of those rare RCU bugs for which a debugger might be helpful. What can you do?

What I have traditionally done is to get partway through figuring out how to make gdb work with rcutorture, then suddenly realize what the bug's root cause must be. At this point, I of course abandon gdb in favor of fixing the bug. As a result, although I have tried to apply gdb to the Linux kernel many times over the past 20 years, I never have actually succeeded in doing so. Now, this is not to say that gdb is useless to Linux-kernel hackers. Far from it! For one thing, the act of trying to use gdb has inspired me to perceive the root cause of a great many bugs, which means that it has served as a great productivity aid. For another thing, I frequently extract Linux-kernel code into a usermode scaffolding and use gdb in that context. And finally, there really are a number of Linux-kernel hackers who make regular use of gdb.

One of these hackers is Omar Sandoval, who happened to mention that he had used gdb to track down a Linux-kernel bug. And without first extracting the code to userspace. I figured that it was time for this old dog to learn a new trick, so I asked Omar how he made this happen.

Omar pointed out that because rcutorture runs in guest OSes, gdb can take advantage of the debugging support provided by qemu. To make this work, you build a kernel with CONFIG_DEBUG_INFO=y (which supplies gdb with additional symbols), provide the nokaslr kernel boot parameter (which prevents kernel address-space randomization from invalidating these symbols), and supply qemu with the -s -S command-line arguments (which causes it to wait for gdb to connect instead of immediately booting the kernel). You then specify the vmlinux file's pathname as the sole command-line argument to gdb. Once you see the (gdb) prompt, the target remote :1234 command will connect to qemu and then the continue command will boot the kernel.

I tried this, and it worked like a charm.

Alternatively, you can now use the shiny new rcutorture --gdb command-line argument in the -rcu tree, which will automatically set up the kernel and qemu, and will print out the required gdb commands, including the path to the newly built vmlinux file.

And yes, I do owe Omar a --drgn command-line argument, which I will supply once he lets me know how to connect drgn to qemu. :-)

In the meantime, the following sections cover a couple of uses I have made of --gdb, mostly to get practice with this approach to Linux-kernel debugging.

Case study 1: locktortureFor example, let's use gdb to investigate a long-standing locktorture hang when running scenario LOCK05:

tools/testing/selftests/rcutorture/bin/kvm.sh --allcpus --torture lock \ --duration 3 --configs LOCK05 --gdb
This will print out the following once the kernel is built and qemu has started:

Waiting for you to attach a debug session, for example: gdb gdb /home/git/linux-rcu/tools/testing/selftests/rcutorture/res/2020.08.27-14.51.45/LOCK05/vmlinux After symbols load and the "(gdb)" prompt appears: target remote :1234 continue
Once you have started gdb and entered the two suggested commands, the kernel will start. You can track its console output by locating its console.log file as described in an earlier post. Or you can use the ps command to dump the qemu command line, looking for the -serial file: command, which is following by the pathname of the file receiving the console output.

Once the kernel is sufficiently hung, that is, more than 15 seconds elapses after the last statistics output line (Writes: Total: 27668769 Max/Min: 27403330/34661 Fail: 0), you can hit control-C at gdb. The usual info threads command will show the CPUs' states, here with the 64-bit hexadecimal addresses abbreviated:

(gdb) info threads Id Target Id Frame * 1 Thread 1 (CPU#0 [running]) stutter_wait (title=0xf... "lock_torture_writer") at kernel/torture.c:615 2 Thread 2 (CPU#1 [running]) 0xf... in stutter_wait ( title=0xf... "lock_torture_writer") at kernel/torture.c:615 3 Thread 3 (CPU#2 [halted ]) default_idle () at arch/x86/kernel/process.c:689 4 Thread 4 (CPU#3 [halted ]) default_idle () at arch/x86/kernel/process.c:689
It is odd that CPUs 0 and 1 are in stutter_wait(), spinning on the global variable stutter_pause_test. Even more odd is that the value of this variable is not zero, as it should be at the end of the test, but rather the value two. After all, all paths out of torture_stutter() should zero this variable.

But maybe torture_stutter() is still stuck in the loop prior to the zeroing of stutter_pause_test. A quick look at torture_stutter_init shows us that the task_struct pointer to the task running torture_stutter lives in stutter_task, which is non-NULL, meaning that this task still lives. One might hope to use sched_show_task(), but this sadly fails with Could not fetch register "fs_base"; remote failure reply 'E14'.

The value of stutter_task.state is zero, which indicates that this task is running. But on what CPU? CPUs 0 and 1 are both spinning in stutter_wait, and the other two CPUs are in the idle loop. So let's look at stutter_task.on_cpu, which is zero, as in not on a CPU. In addition, stutter_task.cpu has the value one, and CPU 1 is definitely running some other task.

It would be good to just be able to print the stack of the blocked task, but it is also worth just rerunning this test, but this time with the locktorture.stutter module parameter set to zero. This test completed successfully, in particular, with no hangs. Given that no other locktorture or rcutorture scenario suffers from similar hangs, perhaps the problem is in rt_mutex_lock() itself. To check this, let's restart the test, but with the default value of the locktorture.stutter module parameter. After letting it hang and interrupting it with control-C, even though it still feels strange to control-C a kernel:

(gdb) print torture_rtmutex $1 = {wait_lock = {raw_lock = {{val = {counter = 0}, {locked = 0 '\000', pending = 0 '\000'}, { locked_pending = 0, tail = 0}}}}, waiters = {rb_root = {rb_node = 0xffffc9000025be50}, rb_leftmost = 0xffffc90000263e50}, owner = 0x1 <fixed_percpu_data+1>}
The owner = 0x1 looks quite strange for a task_struct pointer, but the block comment preceding rt_mutex_set_owner() says that this value is legitimate, and represents one of two transitional states. So maybe it is time for CONFIG_DEBUG_RT_MUTEXES=y, but enabling this Kconfig option produces little additional enlightenment.

However, the torture_rtmutex.waiters field indicates that there really is something waiting on the lock. Of course, it might be that we just happened to catch the lock at this point in time. To check on this, let's add a variable to capture the time of the last lock release. I empirically determined that it is necessary to use WRITE_ONCE() to update this variable in order to prevent the compiler from optimizing it out of existence. Learn from my mistakes!

With the addition of WRITE_ONCE(), the next run showed that the last lock operation was more than three minutes in the past and that the transitional lock state still persisted, which provides strong evidence that this is the result of a race condition in the locking primitive itself. Except that a quick scan of the code didn't immediately identify a race condition. Furthermore, the failure happens even with CONFIG_DEBUG_RT_MUTEXES=y, which disables the lockless fastpaths (or the obvious lockless fastpaths, anyway).

Perhaps this is instead a lost wakeup? This would be fortuitous given that there are rare lost-IPI issues, and having this reproduce so easily on my laptop would be extremely convenient. And adding a bit of debug code to mark_wakeup_next_waiter() and lock_torture_writer() show that there is a task that was awakened, but that never exited from rt_mutex_lock(). And this task is runnable, that is, its ->state value is zero. But it is clearly not running very far! And further instrumentation demonstrates that control is not reaching the __smp_call_single_queue() call from __ttwu_queue_wakelist(). The chase is on!

Except that the problem ended up being in stutter_wait(). As the name suggests, this function controls stuttering, that is, periodically switching between full load and zero load. Such stuttering can expose bugs that a pure full-load stress test would miss.

The stutter_wait() uses adaptive waiting, so that schedule_timeout_interruptible() is used early in each no-load interval, but a tight loop containing cond_resched() is used near the end of the interval. The point of this is to more tightly synchronize the transition from no-load to full load. But the LOCK05 scenario's kernel is built with CONFIG_PREEMPT=y, which causes cond_resched() to be a no-op. In addition, the kthreads doing the write locking lower their priority using set_user_nice(current, MAX_NICE), which appears to be preventing preemption. (We can argue that even MAX_NICE should not indefinitely prevent preemption, but the multi-minute waits that have been observed are for all intents and purposes indefinite.)

The fix (or workaround, as the case might be) is for stutter_wait() to block periodically, thus allowing other tasks to run.

Case study 2: RCU Tasks TraceI designed RCU Tasks Trace for the same grace-period latency that I had designed RCU Tasks for, namely roughly one second. Unfortunately, this proved to be about 40x too slow, so adjustments were called for.

After those reporting the issue kindly verified for me that this was not a case of too-long readers, I used --gdb to check statistics and state. I used rcuscale, which is a member of the rcutorture family designed to measure performance and scalability of the various RCU flavors' grace periods:

tools/testing/selftests/rcutorture/bin/kvm.sh --torture rcuscale --allcpus \ --configs TRACE01 --bootargs "rcuscale.nreaders=0 rcuscale.nwriters=10" \ --trust-make --gdb
Once the (gdb) prompt appears, we connect to qemu, set a break point, and then continue execution:

(gdb) target remote :1234 Remote debugging using :1234 0x000000000000fff0 in exception_stacks () (gdb) b rcu_scale_cleanup Breakpoint 1 at 0xffffffff810d27a0: file kernel/rcu/rcuscale.c, line 505. (gdb) cont Continuing. Remote connection closed (gdb)
Unfortunately, as shown above, this gets us Remote connection closed instead of a breakpoint. Apparently, the Linux kernel does not take kindly to debug exception instructions being inserted into its code. Fortunately, gdb also supplies a hardware breakpoint command:

(gdb) target remote :1234 Remote debugging using :1234 0x000000000000fff0 in exception_stacks () (gdb) hbreak rcu_scale_cleanup Hardware assisted breakpoint 1 at 0xffffffff810d27a0: file kernel/rcu/rcuscale.c, line 505. (gdb) cont Continuing. [Switching to Thread 12] Thread 12 hit Breakpoint 1, rcu_scale_cleanup () at kernel/rcu/rcuscale.c:505 505 {
This works much better, and the various data structures may now be inspected to check the validity of various optimization approaches. Of course, as the optimization effort continued, hand-typing gdb commands became onerous, and was therefore replaced with crude but automatic accumulation and display of relevant statistics.

Of course, Murphy being who he is, the eventual grace-period speedup also caused a few heretofore latent race conditions to be triggered by a few tens of hours of rctorture. These race conditions resulted in rcu_torture_writer() stalls, along with the occasional full-fledged RCU-Tasks-Trace CPU stall warning.

Now, rcutorture does dump out RCU grace-period kthread state when these events occur, but in the case of the rcu_torture_writer() stalls, this state is for vanilla RCU rather than the flavor of RCU under test. Which is an rcutorture bug that will be fixed. But in the meantime, gdb provides a quick workaround by setting a hardware breakpoint on the ftrace_dump() function, which is called when either of these sorts of stalls occur. When the breakpoint triggers, it is easy to manually dump the data pertaining to the grace-period kthread of your choice.

For those who are curious, the race turned out to be an IPI arriving between a pair of stores in rcu_read_unlock_trace() that could leave the corresponding task forever blocking the current RCU Tasks Trace grace period. The solution, as with vanilla RCU in the v3.0 timeframe, is to set the read-side nesting value to a negative number while clearing the .need_qs field indicating that a quiescent state is required. The buggy code is as follows:

if (likely(!READ_ONCE(t->trc_reader_special.s)) || nesting) { // BUG: IPI here sets .need_qs after check!!! WRITE_ONCE(t->trc_reader_nesting, nesting); return; // We assume shallow reader nesting. }
Again, the fix is to set the nesting count to a large negative number, which allows the IPI handler to detect this race and refrain from updating the .need_qs field when the ->trc_reader_nesting field is negative, thus avoiding the grace-period hang:

WRITE_ONCE(t->trc_reader_nesting, INT_MIN); // FIX if (likely(!READ_ONCE(t->trc_reader_special.s)) || nesting) { WRITE_ONCE(t->trc_reader_nesting, nesting); return; // We assume shallow reader nesting. }
This experience of course suggests testing with grace period latencies tuned much more aggressively than they are in production, with an eye to finding additional low-probability race conditions.

To be continued...

James Bottomley: Lessons from the GNOME Patent Troll Incident

Thursday 3rd of September 2020 04:53:19 PM

First, for all the lawyers who are eager to see the Settlement Agreement, here it is. The reason I can do this is that I’ve released software under an OSI approved licence, so I’m covered by the Releases and thus entitled to a copy of the agreement under section 10, but I’m not a party to any of the Covenants so I’m not forbidden from disclosing it.

Analysis of the attack

The Rothschild Modus Operandi is to obtain a fairly bogus patent (in this case, patent 9,936,086), form a limited liability corporation (LLC) that only holds the one patent and then sue a load of companies with vaguely related businesses for infringement. A key element of the attack is to offer a settlement licensing the patent for a sum less than it would cost even to mount an initial defence (usually around US$50k), which is how the Troll makes money: since the cost to file is fairly low, as long as there’s no court appearance, the amount gained is close to US$50k if the target accepts the settlement offer and, since most targets know how much any defence of the patent would cost, they do.

One of the problems for the target is that once the patent is issued by the USPTO, the court must presume it is valid, so any defence that impugns the validity of the patent can’t be decided at summary judgment. In the GNOME case, the sued project, shotwell, predated the filing of the patent by several years, so it should be obvious that even if shotwell did infringe the patent, it would have been prior art which should have prevented the issuing of the patent in the first place. Unfortunately such an obvious problem can’t be used to get the case tossed on summary judgement because it impugns the validity of the patent. Put simply, once the USPTO issues a patent it’s pretty much impossible to defend against accusations of infringement without an expensive trial which makes the settlement for small sums look very tempting.

If the target puts up any sort of fight, Rothschild, knowing the lack of merits to the case, will usually reduce the amount offered for settlement or, in extreme cases, simply drop the lawsuit. The last line of defence is the LLC. If the target finds some way to win damages (as ADS did in 2017) , the only thing on the hook is the LLC with the limited liability shielding Rothschild personally.

How it Played out Against GNOME

This description is somewhat brief, for a more in-depth description see the Medium article by Amanda Brock and Matt Berkowitz.

Rothschild performed the initial attack under the LLC RPI (Rothschild Patent Imaging). GNOME was fortunate enough to receive an offer of Pro Bono representation from Shearman and Sterling and immediately launched a defence fund (expecting that the cost of at least getting into court would be around US$200k, even with pro bono representation). One of its first actions, besides defending the claim was to launch a counterclaim against RPI alleging exceptional practices in bringing the claim. This serves two purposes: firstly, RPI can’t now simply decide to drop the lawsuit, because the counterclaim survives and secondly, by alleging potential misconduct it seeks to pierce the LLC liability shield. GNOME also decided to try to obtain as much as it could for the whole of open source in the settlement.

As it became clear to Rothschild that GNOME wouldn’t just pay up and they would create a potential liability problem in court, the offers of settlement came thick and fast culminating in an offer of a free licence and each side would pay their own costs. However GNOME persisted with the counter claim and insisted they could settle for nothing less than the elimination of the Rothschild patent threat from all of open source. The ultimate agreement reached, as you can read, does just that: gives a perpetual covenant not to sue any project under an OSI approved open source licence for any patent naming Leigh Rothschild as the inventor (i.e. the settlement terms go far beyond the initial patent claim and effectively free all of open source from any future litigation by Rothschild).

Analysis of the Agreement

Although the agreement achieves its aim, to rid all of Open Source of the Rothschild menace, it also contains several clauses which are suboptimal, but which had to be included to get a speedy resolution. In particular, Clause 10 forbids the GNOME foundation or its affiliates from publishing the agreement, which has caused much angst in open source circles about how watertight the agreement actually was. Secondly Clause 11 prohibits GNOME or its affiliates from pursuing any further invalidity challenges to any Rothschild patents leaving Rothschild free to pursue any non open source targets.

Fortunately the effect of clause 10 is now mitigated by me publishing the agreement and the effect of clause 11 by the fact that the Open Invention Network is now pursuing IPR invalidity actions against the Rothschild patents.

Lessons for the Future

The big lesson is that Troll based attacks are a growing threat to the Open Source movement. Even though the Rothschild source may have been neutralized, others may be tempted to follow his MO, so all open source projects have to be prepared for a troll attack.

The first lesson should necessarily be that if you’re in receipt of a Troll attack, tell everyone. As an open source organization you’re not going to be able to settle and you won’t get either pro bono representation or the funds to fight the action unless people know about it.

The second lesson is that the community will rally, especially with financial aid, if you put out a call for help (and remember, you may be looking at legal bills in the six figure range).

The third lesson is always file a counter claim to give you significant leverage over the Troll in settlement negotiations.

And the fourth lesson is always refuse to settle for nothing less than neutralization of the threat to the entirety of open source.

Conclusion

While the lessons above should work if another Rothschild like Troll comes along, it’s by no means guaranteed and the fact that Open Source project don’t have the funding to defend themselves (even if they could raise it from the community) makes them look vulnerable. One thing the entire community could do to mitigate this problem is set up a community defence fund. We did this once before 16 years ago when SCO was threatening to sue Linux users and we could do it again. Knowing there was a deep pot to draw on would certainly make any Rothschild like Troll think twice about the vulnerability of an Open Source project, and may even deter the usual NPE type troll with more resources and better crafted patents.

Finally, it should be noted that this episode demonstrates how broken the patent system still is. The key element Rothschild like trolls require is the presumption of validity of a granted patent. In theory, in the light of the Alice decision, the USPTO should never have granted the patent but it did and once that happened the troll targets have no option than either to pay up the smaller sum requested or expend a larger sum on fighting in court. Perhaps if the USPTO can’t stop the issuing of bogus patents it’s time to remove the presumption of their validity in court … or at least provide some sort of prima facia invalidity test to apply at summary judgment (like the project is older than the patent, perhaps).

Kees Cook: security things in Linux v5.6

Wednesday 2nd of September 2020 11:22:25 PM

Previously: v5.5.

Linux v5.6 was released back in March. Here’s my quick summary of various features that caught my attention:

WireGuard
The widely used WireGuard VPN has been out-of-tree for a very long time. After 3 1/2 years since its initial upstream RFC, Ard Biesheuvel and Jason Donenfeld finished the work getting all the crypto prerequisites sorted out for the v5.5 kernel. For this release, Jason has gotten WireGuard itself landed. It was a twisty road, and I’m grateful to everyone involved for sticking it out and navigating the compromises and alternative solutions.

openat2() syscall and RESOLVE_* flags
Aleksa Sarai has added a number of important path resolution “scoping” options to the kernel’s open() handling, covering things like not walking above a specific point in a path hierarchy (RESOLVE_BENEATH), disabling the resolution of various “magic links” (RESOLVE_NO_MAGICLINKS) in procfs (e.g. /proc/$pid/exe) and other pseudo-filesystems, and treating a given lookup as happening relative to a different root directory (as if it were in a chroot, RESOLVE_IN_ROOT). As part of this, it became clear that there wasn’t a way to correctly extend the existing openat() syscall, so he added openat2() (which is a good example of the efforts being made to codify “Extensible Syscall” arguments). The RESOLVE_* set of flags also cover prior behaviors like RESOLVE_NO_XDEV and RESOLVE_NO_SYMLINKS.

pidfd_getfd() syscall
In the continuing growth of the much-needed pidfd APIs, Sargun Dhillon has added the pidfd_getfd() syscall which is a way to gain access to file descriptors of a process in a race-less way (or when /proc is not mounted). Before, it wasn’t always possible make sure that opening file descriptors via /proc/$pid/fd/$N was actually going to be associated with the correct PID. Much more detail about this has been written up at LWN.

openat() via io_uring
With my “attack surface reduction” hat on, I remain personally suspicious of the io_uring() family of APIs, but I can’t deny their utility for certain kinds of workloads. Being able to pipeline reads and writes without the overhead of actually making syscalls is pretty great for performance. Jens Axboe has added the IORING_OP_OPENAT command so that existing io_urings can open files to be added on the fly to the mapping of available read/write targets of a given io_uring. While LSMs are still happily able to intercept these actions, I remain wary of the growing “syscall multiplexer” that io_uring is becoming. I am, of course, glad to see that it has a comprehensive (if “out of tree”) test suite as part of liburing.

removal of blocking random pool
After making algorithmic changes to obviate separate entropy pools for random numbers, Andy Lutomirski removed the blocking random pool. This simplifies the kernel pRNG code significantly without compromising the userspace interfaces designed to fetch “cryptographically secure” random numbers. To quote Andy, “This series should not break any existing programs. /dev/urandom is unchanged. /dev/random will still block just after booting, but it will block less than it used to.” See LWN for more details on the history and discussion of the series.

arm64 support for on-chip RNG
Mark Brown added support for the future ARMv8.5’s RNG (SYS_RNDR_EL0), which is, from the kernel’s perspective, similar to x86’s RDRAND instruction. This will provide a bootloader-independent way to add entropy to the kernel’s pRNG for early boot randomness (e.g. stack canary values, memory ASLR offsets, etc). Until folks are running on ARMv8.5 systems, they can continue to depend on the bootloader for randomness (via the UEFI RNG interface) on arm64.

arm64 E0PD
Mark Brown added support for the future ARMv8.5’s E0PD feature (TCR_E0PD1), which causes all memory accesses from userspace into kernel space to fault in constant time. This is an attempt to remove any possible timing side-channel signals when probing kernel memory layout from userspace, as an alternative way to protect against Meltdown-style attacks. The expectation is that E0PD would be used instead of the more expensive Kernel Page Table Isolation (KPTI) features on arm64.

powerpc32 VMAP_STACK
Christophe Leroy added VMAP_STACK support to powerpc32, joining x86, arm64, and s390. This helps protect against the various classes of attacks that depend on exhausting the kernel stack in order to collide with neighboring kernel stacks. (Another common target, the sensitive thread_info, had already been moved away from the bottom of the stack by Christophe Leroy in Linux v5.1.)

generic Page Table dumping
Related to RISCV’s work to add page table dumping (via /sys/fs/debug/kernel_page_tables), Steven Price extracted the existing implementations from multiple architectures and created a common page table dumping framework (and then refactored all the other architectures to use it). I’m delighted to have this because I still remember when not having a working page table dumper for ARM delayed me for a while when trying to implement upstream kernel memory protections there. Anything that makes it easier for architectures to get their kernel memory protection working correctly makes me happy.

That’s in for now; let me know if there’s anything you think I missed. Next up: Linux v5.7.

© 2020, Kees Cook. This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 License.

Linux Plumbers Conference: Watch the LPC 2020 Plenary Session

Friday 21st of August 2020 09:56:55 PM

Welcome to LPC 2020! This year we have a record number of attendees,
around 950. We hope you’ll find the conference as engaging and
productive as the ones we had in person for the past 12 years.

Please watch this LPC 2020 welcome message from our Committee Chair, Laura
Abbott, in lieu of our usual plenary session, which also contains useful
information about how this year’s conference will take place.

Enjoy!!

Linux Plumbers Conference: LPC 2020 Is Sold Out

Friday 21st of August 2020 04:33:53 PM

LPC 2020 is sold out. No more tickets are available. We have reached the maximum capacity for our server infrastructure.

Please be considerate, there is no need to contact us asking for tickets, as we are very busy finalizing all the details of the virtual conference.

If you do not have a ticket, you will be able to watch live starting Monday!
Please follow the links here.

Dave Airlie (blogspot): Vallium: a *software* swrast vulkan layer FAQ

Thursday 20th of August 2020 08:08:56 PM

I had some requirements for writing a vulkan software rasterizer within the Mesa project. I took some time to look at the options and realised that just writing a vulkan layer on top of gallium's llvmpipe would be a good answer for this problem. However in doing so I knew people would ask why this wouldn't work for a hardware driver.

tl;dr DO NOT USE VALLIUM OVER A GALLIUM HW DRIVER,

What is vallium?The vallium layer is a gallium frontend. It takes the Vulkan API and roughly translates it into the gallium API.How does it do that?Vulkan is a lowlevel API, it allows the user to allocate memory, create resources, record command buffers amongst other things. When a hw vulkan driver is recording a command buffer, it is putting hw specific commands into it that will be run directly on the GPU. These command buffers are submitted to queues when the app wants to execute them.

Gallium is a context level API, i.e. like OpenGL/D3D10. The user has to create resources and contexts and the driver internally manages command buffers etc. The driver controls internal flushing and queuing of command buffers.
 In order to bridge the gap, the vallium layer abstracts the gallium context into a separate thread of execution. When recording a vulkan command buffer it creates a CPU side command buffer containing an encoding of the Vulkan API. It passes that recorded CPU command buffer to the thread on queue submission. The thread then creates a gallium context, and replays the whole CPU recorded command buffer into the context, one command at a time.That sounds horrible, isn't it slow?Yes.Why doesn't that matter for *software* drivers?Software rasterizers are a very different proposition from an overhead point of view than real hardware. CPU rasterization is pretty heavy on the CPU load, so nearly always 90% of your CPU time will be in the rasterizer and fragment shader. Having some minor CPU overheads around command submission and queuing isn't going to matter in the overall profile of the user application. CPU rasterization is already slow, the Vulkan->gallium translation overhead isn't going to be the reason for making it much slower. For real HW drivers which are meant to record their own command buffers in the GPU domain and submit them direct to the hw, adding in a CPU layer that just copies the command buffer data is a massive overhead and one that can't easily be removed from the vallium layer.
The vallium execution context is also pretty horrible, it has to connect all the state pieces like shaders etc to the gallium context, and disconnect them all at the end of each command buffer. There is only one command submission queue, one context to be used. A lot of hardware exposes more queues etc that this will never model.
I still don't want to write a vulkan driver, give me more reasons.Pipeline barriers:
Pipeline barriers in Vulkan are essential to efficient driver hw usage. They are one of the most difficult to understand and hard to get right pieces of writing a vulkan driver. For a software rasterizer they are also mostly unneeded. When I get a barrier I just completely hardflush the gallium context because I know the sw driver behind it. For a real hardware driver this would be a horrible solution. You spend a lot of time trying to make anything optimal here.Memory allocation:
Vulkan is built around the idea of separate memory allocation and objects binding to those allocations. Gallium is built around object allocation with the memory allocs happening implicitly. I've added some simple memory allocation objects to the gallium API for swrast. These APIs are in no way useful for hw drivers. There is no way to expose memory types or heaps from gallium usefully. The current memory allocation API works for software drivers because I know all they want is an aligned_malloc. There is no decent way to bridge this gap without writing a new gallium API that looks like Vulkan. (in which case just write a vulkan driver already).
Can this make my non-Vulkan capable hw run Vulkan?

No. If the hardware can't do virtual memory properly, or expose features for vulkan this can't be fixed with a software layer that just introduces overhead.


Linux Plumbers Conference: How to Join Virtual LPC 2020

Thursday 20th of August 2020 06:33:25 PM

Only 4 days to the beginning of LPC 2020!

A reminder about how to attend our virtual edition of the Linux Plumbers Conference.

If you are registered, you can participate by joining the Meeting Rooms on our Big Blue Button instance, starting Monday August 24th. You will find a front end showing the schedule for the current day with all the active sessions you can join. If you are having issues, please consult the LPC 2020 Participant Guide.

If you are not registered, you can still watch LPC live streams on YouTube. For how to do this, please refer to this page on our website.

Linux Plumbers Conference: LPC 2020 Schedule Finalized, CfP closed

Wednesday 19th of August 2020 11:12:19 PM

We are very pleased to announce that our final schedule is public!

Please take a look at all the great technical content at this year virtual LPC.
You can view the schedule by main blocks , or by track, or as a complete detailed view.

At this time we are closing the CfPs for all tracks. We have still room for a limited number of Birds of a Feather sessions. If you want to propose one, even during the conference, and the necessary participants are all registered, please send an email to our lpc-contact@lists.linuxplumbersconf.org mailing list.

A heartfelt Thank You to all our Speakers and Microconference Leaders, you all have done an incredible job in making this year’s conference successful, in spite of all the challenges that a remote and distributed conference entails.

See you virtually next week!

Linux Plumbers Conference: Conference Is Sold Out, Watch Live Instead

Monday 17th of August 2020 09:38:27 PM

Hello there, thank you all for the very strong interest in LPC 2020!

We have sold out the last ticket today. We have a lot of attendees and speakers this year for our virtual conference! Almost 1000 registrations!

Do not despair though, because there will be the opportunity to watch the conference live streaming on YouTube. We are still sorting out our channels, but keep an eye on our blog and social media in the next couple of days, where we’ll announce how to watch live.

Thanks, and see you next week!

Linux Plumbers Conference: LPC 2020 T-Shirts and Other Items Are Available

Monday 17th of August 2020 07:29:07 PM

We have received several requests for T-shirts this year.

We have always produced T-shirts for attendees since the early days of LPC, and we don’t want to miss the opportunity to offer them this year too.

Since we are all remote, we have set up a LPC Gift Shop where you can order T-shirts in your favorite colors and sizes. We also have made the designs available if you want to print your own.

Enjoy!

Linux Plumbers Conference: Final passes for sale for Linux Plumbers

Friday 14th of August 2020 04:21:58 PM

We hit our registration cap again and have added a few more passes. The final date for purchasing passes is August 19th at 11:59pm PST. If the passes sell out before then we will not be adding more. Thank you all once again for your enthusiasm and we look forward to seeing you August 24-28!

Michael Kerrisk (manpages): man-pages-5.08 is released

Thursday 13th of August 2020 11:57:39 AM
I've released man-pages-5.08. The release tarball is available on kernel.org. The browsable online pages can be found on man7.org. The Git repository for man-pages is available on kernel.org.

This release resulted from patches, bug reports, reviews, and comments from more than 30 contributors. The release includes more than 190 commits that change around 340 pages.

The most notable of the changes in man-pages-5.08 are the following:
  • Dave Martin added documentation of some arm64-specific prctl() operations to the prctl(2) manual page.
  • Aleksa Sarai added some documentation of magic links to the symlink(7) page.
  • I've added documentation on the use of PID file descriptors with setns() to the setns(2) page.
  • I've made various typographical fixes to many manual pages so that certain characters render better on the terminal or in PDF-format, or both.

Daniel Vetter: Lockdep False Positives, some stories about

Thursday 13th of August 2020 12:00:00 AM

Lockdep is giving false positives are the new the compiler is broken.

— David Airlie (@DaveAirlie) August 8, 2020

Recently we’ve looked a bit at lockdep annotations in the GPU subsystems, and I figured it’s a good opportunity to explain how this all works, and what the tradeoffs are. Creating working locking hierarchies for the kernel isn’t easy, making sure the kernel’s locking validator lockdep is happy and reviewers don’t have their brains explode even more so.

First things first, and the fundamental issue:

Lockdep is about trading false positives against better testing.

The only way to avoid false positives for deadlocks is to only report a deadlock when the kernel actually deadlocked. Which is useless, since the entire point of lockdep is to catch potential deadlock issues before they actually happen. Hence false postives are not avoidable, at least not in theory, to be able to report potential issues before they hang the machine. Read on for what to do in practice.

We need to understand how exactly lockdep trades false positives to better discovery locking inconsistencies. Lockdep makes a few assumptions about how real code does locking in practice:

Invariance of locking rules over time

First assumption baked into lockdep is that the locking rules for a given lock do not change over the lifetime of the lock’s existence. This already throws out a large chunk of perfectly correct locking designs, since state transitions can control how an object is accessed, and therefore how the lock is used. Examples include different rules for creation and destruction, or whether an object is on a specific list (e.g. only a gpu buffer object that’s in the lru can be evicted). It’s not possible to proof automatically that certain code flat out wont ever run together with some other code on the same structure, at least not in generality. Hence this is pretty much a required assumption to make lockdep useful - if every new lock() call could follow new rules there’s nothing to check. Besides realizing that an actual deadlock indeed occured and all is lost already.

And of course getting such state transitions correct, with the guarantee that all the old code will no longer run, is tricky to get right, and very hard on reviewers. It’s a good thing lockdep has problems with such code too.

Common locking rules for the same objects

Second assumption is that all locks initialized by the same code are following the same locking rules. This is achieved by making all lock initializers C macros, which create the corresponding lockdep class as a static variable within the calling function. Again this is pretty much required, since to spot inconsistencies you need as many observations of all the different code path possibilities. Best to share them all between the same object. Also a distinct lockdep class for each individual object would explode the runtime overhead in both memory and cpu cycles.

And again this is good from a code design point too, since having the same data structure and code follow different locking rules for different objects is at best very confusing for reviewers.

Fighting lockdep, badly

Now things go wrong, you have a lockdep splat at your hands, concluded it’s a false positive and go ahead trying to teach lockdep about what’s going on. The first class of annotains are special lock_nested(lock, subclass) functions. Without lockdep nothing in the generated code changes, but it tells lockdep that for this lock acquisition, we’re using a different class to track the observed locking.

This breaks both the time invariance - nothing is stopping you from using different classes for the same lock at different times - and commonality of locking for the same objects. Worse, you can write code which obviously deadlocks, but lockdep will think everything is perfectly fine:

mutex_init(&A); mutex_lock(&A); mutex_lock_nested(&A, SINGLE_DEPTH_NESTING);

This is no good and puts a huge burden on reviewers to carefully check all these places themselves, manually. Exactly the kind of tedious and error prone work lockdep was meant to take over.

Slightly better are the annotations which adjust the lockdep class once, when the object is initialized, using lockdep_set_class() and related functions. This at least does not break time invariance, hence will at least guarantee that lockdep spots the deadlock latest when it happens. It still reduces how much lockdep can connect what’s going on, but occasionally “rewrite the entire subsystem” to resolve a locking inconsistency is just not a reasonable option.

It still means that reviewers always need to remember what the locking rules for all types of different objects behind the same structure are, instead of just one. And then check against every path whether that code needs to work with all of them, or just some, or only one. Again tedious work that really lockdep is supposed to help with. If it’s hard to come by a system where you can easily run the code for the different types of object without rebooting, then lockdep cannot help at all.

All these annotations have in common that they don’t change the code logic, only how lockdep interprets what’s going on.

An even more insideous trick on reviewers and lockdep is to push locking into an asynchronous worker of some sorts. This hides issues because lockdep does not follow dependencies between threads through waiter/wakee relationships like wait_for_completion() and complete(), or through wait queues. There are lockdep annotations for specific dependencies, like in the kernel’s workqueue code when flushing workers or specific work items with flush_work(). Automatic annotations have been attemped with the lockdep cross-release extension, which for various reasons had to be backed out again. Therefore hand-rolled asynchronous code is a great place to create complexity and hide locking issues from both lockdep and reviewers.

Playing to lockdep’s strength

Except when there’s very strong justification for all the complexity, the real fix is to change the locking and make it simpler. Simple enough for lockdep to understand what’s going on, which also makes reviewer’s lifes a lot better. Often this means substantial code rework, but at least in some cases there are useful tricks.

A special kind of annotations are the lock_nest_lock(lock, superlock) family of functions - these tell lockdep that when multiple locks of the same class are acquired, it’s all serialized by the single superlock. Lockdep then validates that the right superlock is indeed held. A great example is mm_take_all_locks(), which as the name implies, takes all locks related to the given mm_struct. In a sense this is not a pure annotation, unlike the ones above, since it requires that the superlock is actually locked. That’s generally the easier to understand scheme than clever sorting of lock acquisition of some sort for reviewers too, not just for lockdep.

A different situation often arises when creating or destroying an object. But at that stage often no other thread has a reference to the object and therefore can take the lock, and the best way to resolve locking inconsistency over the lifetime of an object due to creation and destruction code is to not take any locks at all in these paths. There is nothing to protect against after all!

In all these cases the best option for long term maintainability is to simplify the locking design, not reduce lockdep’s power by reducing the amount of false positives it reports. And that should be the general principle.

tldr; do not fix lockdep false positives, fix your locking

Linux Plumbers Conference: Linux Plumbers Releasing More Passes

Monday 10th of August 2020 06:41:27 PM

After a careful review we have decided to release more passes. We are thrilled with the interest for this first ever online Linux Plumbers. The highlight of Linux Plumbers is the microconferences which are heavily focused on discussion and problem solving. To give the best experience for discussion, we have chosen to use an open source virtual platform that offers video for all participants. The platform recommends not having more than a certain number of people in each room at a time, hence putting a cap on registration to avoid hitting that limit. We do have solutions that will hopefully allow as many people as possible to experience Plumbers. We appreciate your patience and enthusiasm.

Linux Plumbers Conference: Linux Plumbers currently sold out

Saturday 8th of August 2020 05:09:45 PM

Linux Plumbers is currently sold out of regular registration tickets. Although the conference is virtual this year our virtual platform cannot support an unlimited number of attendees, hence the cap on registration. We are currently reviewing our capacity limits to see if we can allow more people to attend without over burdening the virtual platform and potentially preventing discussion. We will make another announcement next week regarding registration.

Linux Plumbers Conference: Toolchain Microconference Accepted into 2020 Linux Plumbers Conference

Friday 7th of August 2020 07:22:50 PM

We are pleased to announce that the Toolchain Microconference has been accepted into the 2020 Linux Plumbers Conference!

The GNU toolchain has direct impact on the development of the Linux kernel and it is imperative that the developers of both ecosystems have an understanding of each other’s needs. Linux Plumbers is the perfect venue for the two communities to interact, and the GNU Toolchain microconference’s purpose is to facilitate that happening.

Last year’s meetup at Linux Plumbers proved that it is critical that the two communities communicate with each other. As a result of last year’s microconference, the GNU toolchain has completed adding support for BPF, in a more flexible and usable way and system call wrappers in glibc were improved. There have been security features derived from the discussions, such as zeroing of registers when entering a function and implicit initialization of atomics.

This year’s topics to be discussed include:

  • How to improve the continual communication between GNU toolchain developers and kernel developers.
  • Creating a standardization on how to compile the kernel that is suitable for all Live kernel patching frameworks.
  • Now that BPF support is in the GNU toolchain, including a new BPF instruction simulator, we’d like some feedback from kernel developers regarding its usability and whether more features are needed.
  • Compact C Type Format (CTF) has been added to the GNU toolchain, and there is work underway to also add BTF support. However some useful features are still missing (like backtraces) and their design needs to be discussed and agreed upon among toolchain and kernel developers.

Come and join us in the discussion about innovating the most efficient and functional toolchain for building the Linux kernel.

We hope to see you there!

More in Tux Machines

Christian Hergert: GtkSourceView Next

Earlier this year I started a branch to track GTK 4 development which is targeted for release by end-of-year. I just merged it which means that our recently released gtksourceview-4-8 branch is going to be our LTS for GTK 3. As you might remember from the previous maintainer, GtkSourceView 4.x is the continuation of the GtkSourceView 3.x API with all the deprecated API removed and a number of API improvements. Currently, GtkSourceView.Next is 5.x targeting the GTK 4.x API. It’s a bit of an unfortunate number clash, but it’s been fine for WebKit so we’ll see how it goes. It’s really important that we start getting solid testing because GtkSourceView is used all over the place and is one of those “must have” dependencies when moving to a new GTK major ABI. Read more

Linux Lite 5.2 Is Now Ready for Testing Based on Ubuntu 20.04.1 LTS

While you’re probably enjoying your Linux Lite 5.0 installation, work has begun on the next major release, Linux Lite 5.2, which will be based on Canonical’s Ubuntu 20.04.1 LTS (Focal Fossa) operating system and the long-term supported Linux 5.4 kernel series. As usual, there are also various improvements and new features. For example, Linux Lite 5.2 will now let users manage the Firewall and Lite Widget settings from the Settings Manager, show laptop battery status in the Lite Widget, as well as to restore the Taskbar and system tray icons to default from the Lite Tweaks utility. Read more

Microsoft is Bringing Edge Browser to Linux in October

At the Ignite 2020, Microsoft announced that the Chromium-based Edge browser will have a Linux preview build in October this year. Read more

Security: Site SSL, Cynet an Ksplice-based Patching

  • Why You Should Use SSL on Your Website

    With the evolution of the internet, security threats have also risen to a great extent. [...] SSL is the digital certificate known as the “Secure Socket Layer” that provides the foundation for stronger security on a website. It acts as a shield and safeguard when sensitive information travels from one place to another between computers/servers. SSL can be defined as trustworthy files that cryptographically form an encrypted link between a browser and a web server. Any information that is sent or received on a page that is not secure can be hacked and intercepted by cyber-criminals and hackers. Important information, such as bank transaction details and personal details become accessible to hackers. A website that is encrypted with SSL binds a secure connection between the web browser and servers to ensure that no third party has access to your information.

  • Cynet Report Details Increase in Cyber Attacks During Pandemic

    Cynet has released a report detailing changes in cyberattacks observed across North America and Europe since the beginning of the COVID-19 pandemic. Cynet compared the number of cyberattacks during the COVID-19 outbreak to the previous three months for several industry sectors and saw increases of more than 20 percent in the areas of finance (up 32%), food production (29%), and retail (23%).

  • Security Patching Made Simple for Linux HPC Instances in Oracle Cloud [Ed: Oracle pushing Ksplice as its Linux selling point]

    The explosion of data in today's computing landscape has fueled the need for even greater security to protect the applications and workloads, and is crucial to an organization's success and competitive advantage. This is equally true when running compute intensive high performance computing (HPC) applications that consume large amounts of data, which are critical to an organization’s business or research endeavors. Oracle Cloud Infrastructure provides a platform that can help keep HPC systems secure and improve the speed and stability of applications. Security patch management is a challenge given the sheer number of instances in HPC clustered environments. Often, HPC environments are left unpatched for long periods of time, leaving systems exposed due to delays caused by complex, time-consuming, and labor-intensive patch management processes. We'll describe three ways in which this is addressed with Oracle Cloud Infrastructure. [...] Ksplice, Autonomous Linux, and the OS Management service are provided for Oracle Cloud Infrastructure customers at no additional cost. Oracle Linux HPC customers on Oracle Cloud Infrastructure enjoy additional benefits including free Oracle Linux Premier Support and price per performance advantages. Additionally, Oracle Linux is 100% application binary compatible with RHEL. This means that RHEL customers on Oracle Cloud Infrastructure can eliminate support fees by easily switching to Oracle Linux. HPC customers who leverage these advanced Linux patching technologies in Oracle Cloud Infrastructure benefit from improved system security, reduced downtime, simplified operations, and cost savings. To learn more about Oracle Cloud patch management options, sign up for an Oracle Cloud Infrastructure account today and take advantage of free cloud credits.