LWN: Comments on "Tracepoints for the VFS?" https://lwn.net/Articles/1017573/ This is a special feed containing comments posted to the individual LWN article titled "Tracepoints for the VFS?". en-us Mon, 29 Sep 2025 04:52:38 +0000 Mon, 29 Sep 2025 04:52:38 +0000 https://www.rssboard.org/rss-specification lwn@lwn.net Method tracepoints https://lwn.net/Articles/1018473/ https://lwn.net/Articles/1018473/ willy <div class="FormattedComment"> I wasn't able to be in the room for this one, sorry.<br> <p> I would suggest that there's a win to be had in adding VFS tracepoints around the calls to filesystems, because we can then remove the tracepoints that individual filesystems have at the entry/exit points of those methods.<br> <p> The problem will be those filesystem authors who insist on retaining their personal tracepoints because they don't want to change their workflow.<br> </div> Tue, 22 Apr 2025 15:52:20 +0000 Enable CONFIG_TRACEPOINTS only? https://lwn.net/Articles/1018183/ https://lwn.net/Articles/1018183/ meyert <div class="FormattedComment"> is there a way to enable CONFIG_TRACEPOINTS only?<br> <p> The commit 5f87f1121895dc09d2d1c1db5f14af6aa4ce3e94 seems to have removed the ability, for some reason.<br> <p> The best I could come up with is:<br> - FTRACE = y<br> - ENABLE_DEFAULT_TRACERS = y (which selects TRACING which selects TRACEPOINTS, but has the side effect of also setting STACKTRACE = y)<br> <p> Bug or feature?<br> </div> Sun, 20 Apr 2025 19:12:34 +0000 Don't we already have VFS tracepoints? https://lwn.net/Articles/1018135/ https://lwn.net/Articles/1018135/ iabervon <div class="FormattedComment"> Presumably, "[T'so] thinks that could be worked around by keeping the tracepoints at the system-call level" means that they're in functions implementing VFS operations that each do the work that one syscall does, but it's easy to interpret as meaning that they'd be in the code that exposes that functionality as syscalls.<br> </div> Sat, 19 Apr 2025 18:23:24 +0000 Don't we already have VFS tracepoints? https://lwn.net/Articles/1018101/ https://lwn.net/Articles/1018101/ kxxt <div class="FormattedComment"> <span class="QuotedText">&gt; What am I missing? Aren't these VFS syscall tracepoints?</span><br> <p> These are syscall tracepoints, definitely not VFS tracepoints.<br> <p> For example, if you make x32 or x86_32 syscalls on x86_64 linux, they won't hit those syscall tracepoints you mentioned. But they will hit the VFS tracepoints (in the future).<br> </div> Sat, 19 Apr 2025 07:16:59 +0000 Don't we already have VFS tracepoints? https://lwn.net/Articles/1018100/ https://lwn.net/Articles/1018100/ alison <div class="FormattedComment"> What am I missing? Aren't these VFS syscall tracepoints?<br> <p> $ sudo bpftrace -l | grep tracepoint:syscalls | grep read <br> tracepoint:syscalls:sys_enter_read<br> tracepoint:syscalls:sys_enter_readahead<br> tracepoint:syscalls:sys_enter_readlink<br> tracepoint:syscalls:sys_enter_readlinkat<br> tracepoint:syscalls:sys_enter_readv<br> . . . <br> <p> $ uname -r<br> 6.12.21-amd64<br> <p> And there are kfuncs too.<br> <p> </div> Sat, 19 Apr 2025 05:13:22 +0000