In D59892#1383479, @imp wrote:More copyright tagging advice.
- Feed Queries
- All Stories
- Search
- Feed Search
- Transactions
- Transaction Logs
Feed All Stories
All Stories
All Stories
Today
Today
In D59892#1383169, @zhaoxiaoqiang007_gmail.com wrote:In D59892#1375242, @brooks wrote:Please don't implement crti and crtn. The _init and _fini functions are obsolete and should not be supported in new architectures. IMO it was an error to have added support to RISC-V.
Is it appropriate to change the clang FreeBSD driver to reflect this ?
I have opened a PR for this： https://github.com/llvm/llvm-project/pull/228689
ngie committed rG72a80a4733b2: poll.2: document that poll is not restarted after signals (authored by _ov4 <ov4@ov4.org>).
@markj : do you have any more comments?
None of my concerns are high enough priority to warrant me blocking my ship-it; the issues can be resolved after the changes are merged.
Approved-by: ngie (co-mentor)
In D61300#1405514, @seuros wrote:In D61300#1405457, @ngie wrote:(Picking a review) what is the intent behind the code refactors? Modularity? Symmetry with acpi?
Maintenance!
Its mentally easier to manage MFCs when we know which files need to stay in sync across branches.
For example, asmcsysctl.c needs to be MFCed every time it changes.T2 support was added in 16 and needs multiple drivers to fully function. So touching one one of it
file automatically exclude it from being MFC.Also having drivers with 4K+ LOC in 1 file is one of the ways we end up accumulating dead code.
kevans committed rGec6629628cbe: linux: Bound native getdents and getdents64 batches (authored by hayzam_gmail.com).
engines support is technically kind of defunct at this point IIRC, but it doesn't hurt to be closer to upstream.
ngie retitled D61294: asmc: split the debug key access into asmc_debug.c from asmc: split the debug key access into asmcdebug.c
ziaee added inline comments to D54734: spi(8) - "-S" option for continuous stream from standard input to bus.
Please add a history section saying when this interface became available.
kib updated the diff for D61253: Implement pthread_mutex_clocklock(3) and pthread_rwlock_clock{wr,rd}lock(3) from POSIX 2024.
Unconditionally call kern_clock_gettime().
Deindent.
Add missed return() in compat code.
aokblast added inline comments to D61253: Implement pthread_mutex_clocklock(3) and pthread_rwlock_clock{wr,rd}lock(3) from POSIX 2024.
markj committed rGedaa6ba7be86: cryptodev: zero key material before freeing it (authored by ojanerif_amd.com).
aokblast added inline comments to D61253: Implement pthread_mutex_clocklock(3) and pthread_rwlock_clock{wr,rd}lock(3) from POSIX 2024.
markj added inline comments to D61253: Implement pthread_mutex_clocklock(3) and pthread_rwlock_clock{wr,rd}lock(3) from POSIX 2024.
In D61215#1404928, @kib wrote:Using pthread locking primitives in signal handler is UB for very good reason. We do not block signals around locks. Then, signal delivery might interrupt either locked region (causing app to recurse on its locks) or the locking code itself (causing libthr to operate on the inconsistent lock state). The pshared component is just a part of the lock state, and it is not async signal safe.
So trying to add the safety for single bit of it does not seem useful.
Yesterday
Yesterday
adrian added a comment to D61330: linux: Move the machine-dependent ioctl bits into per-arch headers.
looks good to me. what do others think?
Updated copyright comments
jhibbits added reviewers for D61330: linux: Move the machine-dependent ioctl bits into per-arch headers: adrian, Linux Emulation.
jhibbits requested review of D61330: linux: Move the machine-dependent ioctl bits into per-arch headers.
Assert that nd_dirfd is not special.
Handle restart for NDINIT_ATVP().