5 A code bounty is a certain financial stimulant to encourage someone to put his or her effort into a project. A code bounty in the DragonFlyBSD project needs to adhere to a set of rules defined hereunder, though these are only preliminary and subject to change. One could for example imagine developers accepting sponsorship on a wholly different basis, or non-financial rewards for completing a bounty project, like free hardware (we love that!).
9 * Bounty Projects are open to anyone that has a reasonable capability of completing the project. Typically it's on first to ask basis, but the sponsor of the bounty may choose differently.
11 * If the developer can not complete the project by the given time period, the bounty agreement is void and no money will be issued for work done. DragonFly developers may extend the time period if it's reasonable and prudent to do so. Upon voiding the agreement, someone may reissue the project again to another developer.
13 * Payment for a project shall be issued in no less than 72 hours after the DragonFly developers has had ten business days to determine if the code is stable and usefulness, no major issues are left unresolved and able for being merged into the project. Payment shall be in either [[Paypal]] (preferred), some money sending company or a bank account transfer.
15 * Upon receiving payment, the project developer shall issue (within 24 hours) a message in the WIKI and to DragonFly development ML that they have received payment.
17 * All code submitted shall be of the same license as DragonFly. Any questions on the licensing issues should be directed to the DragonFly Team.
24 ## In-kernel WireGuard port
28 [WireGuard](https://www.wireguard.com/) is an extremely simple yet fast and modern VPN that utilizes state-of-the-art cryptography.
29 It aims to be faster, simpler, leaner, and more useful than IPsec, while avoiding the massive headache.
30 It intends to be considerably more performant than OpenVPN.
31 WireGuard is designed as a general purpose VPN for running on embedded interfaces and super computers alike, fit for many different circumstances.
32 Initially released for the Linux kernel, it is now cross-platform (Windows, macOS, BSD, iOS, Android) and widely deployable.
33 (copied from [WireGuard official website](https://www.wireguard.com/))
35 We already have the userland implementation in Golang in DPorts (i.e., [`wireguard-go`](https://github.com/DragonFlyBSD/DPorts/tree/master/net/wireguard-go)).
36 It would be better to also gain the in-kernel implementation, like OpenBSD and FreeBSD already did.
40 * Port WireGuard directly from the upstream or from OpenBSD/FreeBSD.
41 * The crypto framework might need an update (from OpenBSD/FreeBSD).
44 The initial port by FreeBSD/Netgate was buggy, so need to investigate before using that code.
45 See: [Buffer overruns, license violations, and bad code: FreeBSD 13's close call](https://arstechnica.com/gadgets/2021/03/buffer-overruns-license-violations-and-bad-code-freebsd-13s-close-call/)
49 * WireGuard port in DragonFly kernel, as well as the userland tools.
51 * The port should be ideally reviewed/audited by the WireGuard project.
52 I (aly@) believe the author (Jason A. Donenfeld) would like to help.
56 Created at: 2021-08-12
82 ## Bring in support for USB webcams
86 Bring in `uvideo` from NetBSD or OpenBSD, and the dependent `video` driver. If you think that webcam support in FreeBSD is superior, feels free to bring in `cuse` etc.
88 Final result should be that USB webcams are usable from chrome or firefox for the purpose of video conferencing (e.g. via jitsi.org). Personally I only need support for low-quality webcams, no need for 4k :).
92 Bounty created at: 2021-08-08
122 <td>TOTAL AMOUNT</td>
123 <td>200 USD + 100 GBP</td>
129 ## Port uhidd HID daemon from FreeBSD
133 uhidd is a userland driver/daemon for USB HID devices like mouse, keyboard etc.
134 It is based on libusb20 and currently only works for FreeBSD 8.x and above.
136 The uhidd driver would allow USB gaming controllers and joysticks to be used on DragonFly BSD.
137 uhidd also provide support for multimedia keys found on USB keyboards.
139 (Derived from the [FreeBSD Wiki: uhidd](https://wiki.freebsd.org/uhidd))
141 ### Technical details
143 * This mainly involves porting uhidd and related items from FreeBSD.
147 * Being able to use USB game controllers and joysticks on DragonFly.
151 Bounty created at: 2020-02-17
171 <td>TOTAL AMOUNT</td>
178 ## Extend vkernel(7) to run on any POSIX OS
182 * vkernel(7) allows for running DragonFly kernels in userland. Still, it requires a DragonFly operating system.
184 * Similar to NetBSD's rumpkernel, allow the vkernel application to be independent of the operating system. Using a cross-compiler,
185 this would allow us to run a DragonFly kernel for instance on Linux or NetBSD (I know that running it within a virtual machine would be faster...).
187 ### Technical details
189 * This mainly involves rewriting parts of the page-table handling code and interrupt-related code.
193 * Being able to run a vkernel binary on any POSIX OS.
197 Bounty created at: 2020-02-12
217 <td>TOTAL AMOUNT</td>
228 * Make Bluetooth work on DragonFly.
230 ### Technical details
232 We already have bluetooth support in the kernel which is based on netgraph6 (AFAIK).
233 It is disabled by default and is probably out of date.
235 We also have a partial and old netbt port from NetBSD...
239 * Research the current state of Bluetooth in DragonFly.
240 * Bluetooth is enabled by default when building the kernel/world.
241 * Bring in updates or re-port from FreeBSD/NetBSD.
242 * Bluetooth keyboard / mice is working.
243 * Bluetooth headphone is working.
247 Bounty created at: 2020-02-12
277 <td>TOTAL AMOUNT</td>
284 ## Full Rust libc FFI binding support
288 * Port Rust FFI libc bindings, in full to DragonFly.
290 ### Technical details
292 There has intermittently been some support on this and builds and tests have gone through periods of passing and failing, but a comprehensive buildout of full support of all features would make more advanced Rust development, especially as a viable web server backend more possible.
294 The Github URL is https://github.com/rust-lang/libc
298 * Ensure that all features available in our libc have been ported to working in the Rust FFI lib
299 * Ensure that the code builds
300 * Ensure that all tests pass
301 * Ensure at least a semi automated way for the upstream team to be able to continue validation by some sort of CI/CD system so that we don't have a large support divergence
331 <td>TOTAL AMOUNT</td>
339 ## More Modern Thunderbolt Support
343 * Integrate Thunderbolt Devices
345 ### Technical Details
366 <td>Mike Carroll</td>
371 <td>TOTAL AMOUNT</td>
378 ## OpenPower9 Support
382 * Port DragonFly to OpenPower 9 CPUs
384 ### Technical Details
405 <td>Mike Carroll</td>
410 <td>TOTAL AMOUNT</td>
421 * Port DragonFly to Aarch64
423 ### Technical details
449 <td>Mike Carroll</td>
454 <td>TOTAL AMOUNT</td>
455 <td>100 EUR + 150 USD</td>
461 ## UTF-8 support in Console
465 Adding full UTF-8 support in console.
467 ### Technical details
503 <td>TOTAL AMOUNT</td>
504 <td>35 EUR + 150 USD</td>
509 ## HAMMER compression
513 (shamelessly copied from gsoc2010 projects page)
515 * Compress blocks as they get written to disk.
516 * Only file data (rec_type == DATA) should be compressed, not meta-data.
517 * the CRC should be that of the uncompressed data.
518 * ideally you'd need to associate the uncompressed data with the buffer cache buffer somehow, so that decompression is only performed once.
519 * compression could be turned on a per-file or per-pfs basis.
520 * gzip compression would be just fine at first; lzo or lzjb might be preferable.
523 HAMMER2 has replaced HAMMER to be the default filesystem on DragonFly and is under active development and maintenance. HAMMER2 also has compression support.
525 ### Technical details
527 (shamelessly copied from gsoc2010 projects page)
529 Doing compression would require flagging the data record as being compressed and also require double-buffering since the buffer cache buffer associated with the uncompressed data might have holes in it and otherwise referenced by user programs and cannot serve as a buffer for in-place compression or decompression.
531 The direct read / direct write mechanic would almost certainly have to be disabled for compressed buffers and the small-data zone would probably have to be used (the large-data zone is designed only for use with 16K or 64K buffers).
565 <td>TOTAL AMOUNT</td>
566 <td>50 EUR + 150 USD</td>
572 ## Native NTFS support
576 (shamelessly copied from gsoc2010 projects page)
578 * Port Apple's native NTFS driver
580 ### Technical details
582 (shamelessly copied from gsoc2010 projects page)
584 Apple's NTFS native driver could be ported to DragonFly BSD.
586 Latest code as of now is:
587 http://opensource.apple.com/source/ntfs/ntfs-84.40.1/kext/
611 <td>TOTAL AMOUNT</td>
618 ## Card reader driver
622 As swildner wrote on 18th Septmber 2017: we need a new driver for a card reader.
624 RTS5209 PCI Express Card Reader
644 <td>TOTAL AMOUNT</td>
654 ## Port valgrind to DragonFlyBSD [TAKEN 2021-03-30]
658 (shamelessly copied from gsoc2010 projects page)
660 Valgrind is a very useful tool on a system like DragonFly that's under heavy development. Ideally, we would want the port to be usable with vkernel processes, thus enabling complex checking of the core kernel code.
662 ### Technical details
668 (shamelessly copied from gsoc2010 projects page)
670 The goal of this project is to port valgrind (3.5.0+) to the DragonFlyBSD platform so that at least the memcheck tool runs sufficiently well to be useful. Also an update for the pkgsrc package (devel/valgrind) should be made.
680 Repository: [[Github|https://github.com/1338-1/valgrind-dragonfly]]
711 <td>TOTAL AMOUNT</td>
712 <td>150 EUR + 150 USD (paid)</td>
717 ## Hypervisor for DragonFly/x86_64
721 * Port nvmm (NetBSD) or bhyve (FreeBSD) to DragonFly.
723 ### Technical details
725 Employ Intel VT-x and AMD SVM to provide fast virtualization support for running various guest OSes.
729 * Hardware-accelerated virtualization to run guest operating systems like Linux or Windows 10 on DragonFly.
730 * Support both Intel and AMD CPUs.
734 Created at: 2020-02-12<br>
735 Completed at: 2021-08-08
743 Completed by: Aaron LI (aly@)<br>
745 Previous Party: Kamil Rytarowski (kamil at netbsd dot org)<br>
746 Previous Repository: https://github.com/Moritz-Systems/DragonFlyBSD
750 * Ported NVMM from NetBSD-current (as of 2021-06-25), plus various improvements/fixes from maxv's repo.
751 * Support both Intel VMX and AMD SVM on x86_64 CPUs.
753 - nvmm(4): kernel module, including both the MI frontend and MD backends
754 - libnvmm(3): hypervisor API interacting with nvmm(4), e.g., used by QEMU
755 - nvmmctl(8): utility to inspect NVMM status (still premilinary)
756 - testcases/libnvmm: libnvmm test cases
757 - test/nvmm: examples
758 * Merged into master (6.1) on 2021-07-20.
759 * QEMU (6.0.0) has been updated to employ NVMM on DragonFly and available in [DPorts: emulators/qemu60](https://github.com/DragonFlyBSD/DPorts/tree/master/emulators/qemu60).
760 * HowTo document: [Virtualization: NVMM Hypervisor](/docs/docs/howtos/nvmm)
772 <td>500 USD (paid)</td>
777 <td>175 USD (paid)</td>
782 <td>500 USD (paid)</td>
787 <td>500 CNY (myself ;))</td>
791 <td>TOTAL AMOUNT</td>
792 <td>1175 USD + 500 CNY</td>
798 ## Port autofs from FreeBSD
802 Autofs allows dynamic filesystem mapping to render filesystems accessible, without keeping them mounted. It is designed as a modern, simpler, cleanroom reimplementation of amd.
822 <td>50 USD (paid)</td>
826 <td>TOTAL AMOUNT</td>
837 * Completely rip out our USB stack and replace it with FreeBSD's new usb4bsd (or whatever the FreeBSD-current USB stack is called nowadays).
838 * Note that some work was done here already by polachok (http://gitweb.dragonflybsd.org/~polachok/dragonfly.git/shortlog/refs/heads/usb2) but the state is unknown.
840 ### Technical details
862 <td>300 USD (paid)</td>
867 <td>100 USD (paid)</td>
872 <td>50 EUR (paid)</td>
877 <td>50 EUR (paid)</td>
881 <td>TOTAL AMOUNT</td>
882 <td>100 EUR + 400 USD</td>