delamon

4 ms·

1.
by delamon 6d ago
Yes, I mean the code in the drop implementation.
2.
by delamon 7d ago
I don't quite get it. 5-byte utf-8 encoding gets extra 5 bits compared to 4 byte, and 6-byte gets extra 10 bits. If you were thinking about bits in leading byte, then yes, you are losing one bit for every extra trailing byte, but you a
3.
by delamon 7d ago
Yes, that would work. But you would need to carry pointer to allocator inside box and it is extra code to run for every object.
4.
by delamon 7d ago
6. modern nvme drive, preferrably pcie gen5 that can push >14GiB/sec
5.
by delamon 7d ago
They don't play nicely with arena allocators. And arenas is what you reach for if you have clear lifetime bounds: e.g. a single request with arena never de-allocates individual objects, nukes arena when done. That gives you an easy ver
6.
by delamon 7d ago
We still have about 85% of codepoint space unused. Hopefully, by the time it becomes a problem, UTF-16 will be long dead
7.
by delamon 9d ago
I have following lines "\e[A": history-search-backward "\e[B": history-search-forward in my .inputrc for years. I think, this is how command line worked in tcsh. It changes unconditional prev/next to prefix
8.
by delamon 22d ago
> Turkey for example has multiple frequencies under 5Ghz WiFi that is banned, hence consumer don't get to use 5Ghz WiFi. Are you sure? According to wiki, there are plenty allowed 5GHz channels in Turkey.
9.
by delamon 27d ago
Objects belonging to multiple double-linked lists at the same time. Easily done with intrusive lists. Safe rust would require Rc/Arc: penalty both on memory usage and cpu time.
10.
by delamon 28d ago
The code is sloppy. They check existance of kdialog binary using full path; next step they rely on PATH search by shell. If would've been much safer to just do execve directly.
11.
by delamon 1mo ago
My amd thinkpad idles at 3w, watching youtube is about 4.5-5w.
12.
by delamon 1mo ago
Yes, but if you want pus truly wide events (e.g. >64KiB), then it migth push syslog implementation from happy/well-throded path.
13.
by delamon 1mo ago
> There can be overload, but the backpressure should propagate downstream and senders and intermediaries should buffer locally if needed. You're only delaying the inevitable. Even with local buffering, you can arrive at a point wher
14.
by delamon 2mo ago
You surely meant capacitors, not transistors
15.
by delamon 2mo ago
How so? PGO might indicate that most of the time arguments are in bounds, but that's not enough. It has to be a proof with 100% certainty before bound checks can be removed
16.
by delamon 2mo ago
Or even use acl(5), which is simplier than making arbitrary groups.
17.
by delamon 3mo ago
You can, in fact, give compiler a hint wether the branch will be predictable or not. E.g. clang has __builtin_unpredictable().
18.
Claude bug report: Cross-session credential leakage(github.com)
4 points by delamon 3mo ago|0 comments
19.
by delamon 3mo ago
He only gets 4-5Gbps in the other direction.
20.
by delamon 4mo ago
Inderect call cost is a few cycles, if predicted. Now, you can argue, that it may be mispredicted and misprediction would cost about 20-30 cycles. But if it is mispredicted, then you are not calling into allocator often enough. And if you d
21.
Tanenbaum: Can We Make Operating Systems Reliable and Secure? (2006) [pdf](cs.vu.nl)
2 points by delamon 4mo ago|0 comments
22.
by delamon 4mo ago
can't help myself: how common to see code like https://github.com/oven-sh/bun/blob/main/src/bun_core/string... instead of simple ilog2()? I guess, ut somewhat explains how they got 700k+ l
23.
by delamon 5mo ago
yes
24.
by delamon 5mo ago
I assume you mean container, not VM. But yes, container makes it harder.
25.
by delamon 5mo ago
The attack gives you ability to overwrite any cached page. So you don't need to be root to "edit" /etc/passwd.
26.
by delamon 5mo ago
setuid binaries are not the only way to get root. E.g. one can change /etc/crontab or /etc/passwd. Or add trojan to /bin/ls and wait until admin type 'ls'
27.
by delamon 5mo ago
Apparently it is both...
28.
by delamon 5mo ago
If you "forget" to mark COW memory pointer as no-write, the net effect would be same, would it not? If I'm reading the diff correctly, the problem was that code missed to mark some pages as shared (aka no-write).
29.
by delamon 5mo ago
How would've memory safety helped here?
30.
by delamon 5mo ago
Another data point: AMD T14 G4 with s2idle draws about 0.4% per hour of sleep.

More ›