Backport PRs to release-0.20 #1260
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Projects such as llvm require code to conform to rv32i2p1_m2p0 and no
longer support rv32i2p0_m2p0. This breaks compilation when these files
are linked by llvm giving errors like so:
```
error: invalid arch name 'rv32i2p0_m2p0', unsupported version number 2.0 for extension 'i'
.attribute 5, "rv32i2p0_m2p0"
^
```
This change relaxes the architecture versioning by dropping the version
numbers for each feature and using `rv32im` instead. By doing so, we
will be able to use newer versions of rust and llvm to generate the
guest code.
`sys_log` doesn't insert a newline and makes the terminal output messy. Add a newline for cleaner output. --------- Co-authored-by: Frank Laub <flaub@risc0.com>
As described in #1243, our use of the `jal` instruction in our entrypoint breaks when the jump distance becomes too far. This PR changes the instruction to the `call` psuedo-instruction to allow the assembler flexibility to solve this issue and implement a far jump. Co-authored-by: @weikengchen Resolves: #1243 --------- Co-authored-by: Frank Laub <frank@risczero.com> Co-authored-by: Frank Laub <flaub@risc0.com>
Rust naming guidelines recommend C-GETTER conventions for "getter function". This PR renames `env::get_cycle_count` to `env::cycle_count` to follow these guidelines https://rust-lang.github.io/api-guidelines/naming.html#getter-names-follow-rust-convention-c-getter --------- Co-authored-by: Frank Laub <frank@risczero.com>
|
The latest updates on your projects. Learn more about Vercel for Git ↗︎
|
Co-authored-by: Frank Laub <flaub@risc0.com>
This PR changes the Bonsai SDK to incorporate composition: - [x] Add the option to include a list of UUID representing receipts as part of the `create_session` payload - [x] Add the `upload_receipt` route - [x] Add the test to both new additions - [x] Update alpha_async crate - [x] Update Bonsai SDK crate - [x] Update ethereum crate - [x] Update rest-api-mock-crate
This PR adds the `resolve` method to the `Client` API such that a succinct receipt can be generated by repeated calls to the lift, join, and resolve methods on `Client`.
Adds a JSON Web Token validator to demonstrate validating RS256 signed tokens in the guest. --------- Co-authored-by: Frank Laub <flaub@risc0.com>
Composition currently does not work in client/server mode because receipts are not sent from client to server. This PR adds the required plumbing to send the assumptions from client to server. Resolves: #1236
Happy New Year! I've omitted the circuit from this PR but I will make a separate update after updating WIP and bootstrapping. It looks like tests in doc-test/version-0.18 is failing due to the `risc0-zkvm-methods` crate failing to build. I'm going to omit changes to that directory for the time being.
I spent a momment to look into the minimal guest programs under `std`
and `no_std` configurations, and found that in the `std` case we can
drop the non-standard lines we use today. This PR changes the minimal
`std` guest in our examples and templates to the following:
```rust
use risc0_zkvm::guest::env;
fn main() {
env::commit("hello");
}
```
By contrast, here was the `std` guest we were using before:
```rust
use risc0_zkvm::guest::env;
risc0_zkvm::guest::entry!(main);
fn main() {
env::commit(&0xf00u64);
}
```
In the case of `no_std`, I attempted to improve upon the current minimal
guest, and did find an alternative. What I discovered is that removing
`#![mo_main]` means we will need to define a `start` langauge symbol. I
found that the standard way to do that is with the `start` feature,
resulting in the following guest.
```rust
use risc0_zkvm::guest::env;
fn main(_: isize, _: *const *const u8) -> isize {
env::commit("hello");
0
}
```
Comapred with what we have, this seems less readable. As a result, I am
not suggesting we change the minimal prgram from `no_std`. It will
remain as:
```rust
use risc0_zkvm::guest::env;
risc0_zkvm::guest::entry!(main);
fn main() {
env::commit(&0xf00u64);
}
```
This PR additionally removes the `pub` visibility modifier from all
`main` functions except `r0vm`, as it is not needed.
---------
Co-authored-by: Frank Laub <flaub@risc0.com>
The changes the MemorySet trace event to use a vector of bytes instead of a u32. By doing so, TraceEvent can capture bytes written a call to `store_region` using one trace event. Fixes: #1194 --------- Co-authored-by: Frank Laub <flaub@risc0.com>
We had a bug previously where execution statistics were reported with inaccurate values. This PR seeks to solve this issue.
This PR adds a basic example of composition. Building upon the proof of known factors, this example exponentiates a number mod a composite with known factors. This takes inspiration from RSA encryption, and is kind of a stripped down (and insecure) version of verifiable encryption such that the message is provably decrypt-able by some party. Resolves: #1299 --------- Co-authored-by: Frank Laub <flaub@risc0.com>
This change removes SYS_READ_AVAIL because it's not POSIX compliant and introduces some difficulty in cases where the number of available bytes might be undefined or non-deterministic concerning guest execution. This change removes the SYS_READ_AVAIL support from the host and guest. The guest syscall libraries used SYS_READ_AVAIL to create a buffer of the correct size. Due to its removal, we're allowing cases where EOF is encountered earlier and a shorter read than requested is possible through the guest syscall handlers provided. The underlying handler in the prover has changed to reflect this reality by removing asserts. This is based on @austinabell's solution with minor adjustments. Summary of guest-side syscall handler changes: The previous implementation relied on the number of bytes returned by SYS_READ_AVAIL or has been requested. Logic in SYS_READ has been changed to use the actual amount of bytes read from the syscall. Also, if the guest requests more bytes than available, it short-circuits the requests and returns the amount of read bytes. Summary of host changes: Changed the host to allow the guest to request more bytes than available. In this case, the host will hit an EOF and return the buffer to the caller while still allowing the last unaligned word to be read. Fixes: #1254 --------- Co-authored-by: Austin Abell <austinabell8@gmail.com>
Although very clever, the term "corroborating receipt" was introducing uncessesary confusion. This removes the use of the word "corroborating" with reference to receipts that attest the the validity of assumptions.
…example version (#1320) Co-authored-by: Frank Laub <flaub@risc0.com>
) This PR adds terminology entries to the website for terms/concepts related to composition. Additionally is adds a bit more documentation to types and methods related to composition, and additionally refactors the `ExecutorEnv::add_assumption` method to accept `impl Into<Assumption>` which allows the user to pass in a receipt without any transformation. --------- Co-authored-by: Tim Zerrell <tim.zerrell@risczero.com>
Groth16 prover * Add workflow for generation of groth16 proof of STARK verification. * Fix xtask bootstrap to break cyclic dependency --------- Co-authored-by: Frank Laub <flaub@risc0.com>
…ns (#1334) Newer toolchains have a more detailed panic message in the guest including the line number of the panic. Without this change the 1.75.0 toolchain will fail the `guest_error_forwarding` test like so: ``` running 1 test test host::api::tests::guest_error_forwarding - should panic ... FAILED failures: ---- host::api::tests::guest_error_forwarding stdout ---- thread 'host::api::tests::guest_error_forwarding' panicked at risc0/zkvm/src/host/api/tests.rs:146:22: called `Result::unwrap()` on an `Err` value: Guest panicked: panicked at src/bin/multi_test.rs:99:13: MultiTestSpec::Panic invoked note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace note: panic did not contain expected string panic message: `"called `Result::unwrap()` on an `Err` value: Guest panicked: panicked at src/bin/multi_test.rs:99:13:\nMultiTestSpec::Panic invoked"`, expected substring: `"Guest panicked: panicked at 'MultiTestSpec::Panic invoked'"` ``` Update the expected panic message so it works for older and newer toolchains.
Currently this will read past the first digest until the block is full (2*digest). In the case that the second digest is directly after the first in memory, this read is just redundant. In the case that they are not, when reading just one, there are unnecessary reads done past the first digest in memory. I might be missing something here, but checking manually and running some basic tests work, so just opening this to see if and what changes/breaks to see if I am missing something. --------- Co-authored-by: Frank Laub <flaub@risc0.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
No description provided.