- User Since
- Jul 4 2018, 7:23 PM (144 w, 6 d)
Fri, Apr 9
Sat, Apr 3
Thu, Apr 1
intptr_t isn't sufficient for hybrid CHERI though, so we'd still need the diff to make it intcap_t downstream...
Would (u)intptr_t make more sense so it's faster on CHERI? Doesn't make a difference upstream but does to us.
Tue, Mar 30
Sat, Mar 27
Mon, Mar 22
Sun, Mar 21
NB: Tests submitted upstream at https://gnats.netbsd.org/cgi-bin/query-pr-single.pl?number=56066
Sat, Mar 20
Fri, Mar 19
Mar 14 2021
Mar 12 2021
This warning very rarely useful (it's a hint and not mandatory) and it
Mar 8 2021
Mar 4 2021
Mar 1 2021
Feb 23 2021
Feb 22 2021
Feb 20 2021
Feb 17 2021
Feb 16 2021
MIPS should work fine too, and PowerPC64 ELFv1 really needs a comment saying why TLS is broken (and what it would take to remove this hack), though I'm unconvinced it is given that e.g. jemalloc doesn't die a horrible death.
Feb 10 2021
This is because libm has many functions which are implemented in terms of other math functions. For instance, many functions which handle complex arguments are implemented in terms of calling the non-complex variants. E.g. csinh() calls sinh(), cosh(), sin() and cos(). In all these cases, unexpected outcomes could occur if the compiler deciders to shortcut the logic by inserting its own builtins at those call sites.
Doing so for tests makes sense, but why libm itself?
If these are preprocessor-time constants you could do #if MTHCA_ICM_PAGE_SIZE < PAGE_SIZE instead.
Needs a proper commit message but this should be fine to land now?
Feb 9 2021
Feb 8 2021
I feel like these shouldn't be compiled in by default, as 99.9999% of the time you're not running with TSan and it will just hurt performance.
memcpy is signal safe since POSIX.1-2016, so that is a bug in TSan.
Feb 5 2021
Hm. Perhaps this should be word_size (or similar) and __SIZEOF_LONG__ rather than pointer_size and __SIZEOF_POINTER__ so there's less churn for CHERI?
Feb 1 2021
Jan 30 2021
Jan 28 2021
Jan 27 2021
Jan 26 2021
Manually pasted reduced-context diff to work around Phabricator's UTF-8-centric views
Hmph. Phabricator does not support ISO-8859-1 (their docs say so explicitly) so it balks on these files and treats them as binary. That's going to be fun.
Rebase against shiny new repo
Jan 25 2021
Jan 24 2021
Jan 23 2021
Jan 22 2021
"freezes when building failures on macOS" in the commit message needs fixing
Jan 21 2021
Rebase and address review comments
Don't know if I'm supposed to be able to push with a Reviewed by: + Approved by: debdrup, or if something just hates me, but I get "FATAL: VREF/APPROVERS-CHECK/doc: helper program exit status 65280", so could you please commit this for me?
Fix unintended whitespace diff
The documentation for these says "When set to a nonempty string", and the GNU world does the same (e.g. see LD_TRACE_LOADED_OBJECTS=1 /bin/echo on a GNU/Linux system), so the strcmp shouldn't be added.