From: "Thomas Weißschuh" <linux@weissschuh.net>
To: Johannes Berg <johannes@sipsolutions.net>
Cc: Tiwei Bie <tiwei.bie@linux.dev>,
richard@nod.at, anton.ivanov@cambridgegreys.com,
benjamin@sipsolutions.net, arnd@arndb.de,
linux-um@lists.infradead.org, linux-kernel@vger.kernel.org,
tiwei.btw@antgroup.com
Subject: Re: [PATCH v2 03/10] um: vdso: Implement __vdso_getcpu() via syscall
Date: Mon, 22 Sep 2025 18:04:46 +0200 [thread overview]
Message-ID: <61ae09df-d65b-4c9d-a0c1-7de915246590@t-8ch.de> (raw)
In-Reply-To: <366bb558c3fd23b9a80008d923f29ed0234e17b9.camel@sipsolutions.net>
On 2025-09-22 17:14:18+0200, Johannes Berg wrote:
> On Mon, 2025-09-22 at 16:01 +0200, Thomas Weißschuh wrote:
> > Right now it does not provide any advantage over a regular syscall.
> > Essentially it is just overhead. That said, if you do want to make a
> > real vDSO out of it, I'd be happy to help in that.
>
> I don't know if I'd say "just overhead" - depends on which path is more
> optimised in a typical libc implementation? I'd basically think it's
> identical, no? You either link to the vDSO, or a __weak same function in
> the libc?
The code also needs to be built and maintained. AFAIK __weak is only for
the compile-time linker. The vDSO call will be an indirect call.
> > > I mean ... on the one hand, sure, it doesn't really do much after this,
> > > but OTOH it lets userspace actually use that path? So might be useful.
> >
> > What advantage does userspace have from it?
>
> Right now, none? But it's easier to play with if you have the
> infrastructure, and I'm not convinced there's a _disadvantage_?
So far that hasn't happened. The disadvantages are the ones from above,
nothing critical. But of course it is your subsystem and your call to make.
> > > > Also the functionality to map the host vDSO and vsyscall page into UML
> > > > userspace looks very weird and error-prone. Maybe it can also go away.
> > >
> > > Surely host vDSO etc. is never mapped into UML userspace and never is,
> > > not sure what you're thinking of, but clearly that's wrong as written.
> >
> > This is how I understand the 32bit implementation using
> > ARCH_REUSE_HOST_VSYSCALL_AREA and NEW_AUX_ENT(AT_SYSINFO_EHDR, vsyscall_ehdr)
> > where vsyscall_ehdr comes from the hosts getauxval(AT_SYSINFO_EHDR).
>
> Huh, hm, yeah I forgot about that ... 32-bit. Yeah, agree we should just
> kill that. I'm not even sure it works with the host kernel trapping
> there? Oh well.
Ack, do you want me to send a patch? This was my real gripe with the UM
vDSO. I want to enable time namespaces for all architectures but these
need to be handled in the vDSO properly. For the 64-bit stub vDSO it's
not a problem as the syscalls will work correctly.
But the interaction with the weird 32-bit logic on the other hand...
Thomas
next prev parent reply other threads:[~2025-09-22 16:04 UTC|newest]
Thread overview: 43+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-08-10 5:51 [PATCH v2 00/10] um: Add SMP support Tiwei Bie
2025-08-10 5:51 ` [PATCH v2 01/10] um: Stop tracking virtual CPUs via mm_cpumask() Tiwei Bie
2025-08-10 5:51 ` [PATCH v2 02/10] um: Remove unused cpu_data and current_cpu_data macros Tiwei Bie
2025-08-10 5:51 ` [PATCH v2 03/10] um: vdso: Implement __vdso_getcpu() via syscall Tiwei Bie
2025-09-10 11:59 ` Johannes Berg
2025-09-11 4:29 ` Tiwei Bie
2025-09-21 20:00 ` Thomas Weißschuh
2025-09-22 4:50 ` Tiwei Bie
2025-09-22 12:05 ` Thomas Weißschuh
2025-09-22 12:12 ` Johannes Berg
2025-09-22 14:01 ` Thomas Weißschuh
2025-09-22 15:14 ` Johannes Berg
2025-09-22 16:04 ` Thomas Weißschuh [this message]
2025-09-22 17:07 ` Johannes Berg
2025-09-25 17:08 ` Thomas Weißschuh
2025-10-21 13:20 ` Johannes Berg
2025-08-10 5:51 ` [PATCH v2 04/10] um: Turn signals_* into thread-local variables Tiwei Bie
2025-09-10 12:15 ` Johannes Berg
2025-09-11 4:34 ` Tiwei Bie
2025-09-11 7:37 ` Benjamin Berg
2025-09-11 8:06 ` Benjamin Berg
2025-09-12 0:30 ` Tiwei Bie
2025-09-12 7:58 ` Benjamin Berg
2025-09-12 13:27 ` Tiwei Bie
2025-09-11 9:44 ` Johannes Berg
2025-09-11 10:35 ` Benjamin Berg
2025-08-10 5:51 ` [PATCH v2 05/10] um: Determine sleep based on need_resched() Tiwei Bie
2025-09-10 12:10 ` Johannes Berg
2025-09-11 4:39 ` Tiwei Bie
2025-09-11 6:59 ` Johannes Berg
2025-09-12 0:59 ` Tiwei Bie
2025-09-11 9:27 ` Johannes Berg
2025-09-12 0:54 ` Tiwei Bie
2025-08-10 5:51 ` [PATCH v2 06/10] um: Define timers on a per-CPU basis Tiwei Bie
2025-08-10 9:49 ` kernel test robot
2025-08-10 5:51 ` [PATCH v2 07/10] um: Remove unused ipi_pipe field from cpuinfo_um Tiwei Bie
2025-08-10 5:51 ` [PATCH v2 08/10] um: Add initial SMP support Tiwei Bie
2025-09-11 9:32 ` Johannes Berg
2025-09-12 0:45 ` Tiwei Bie
2025-09-12 7:58 ` Johannes Berg
2025-08-10 5:51 ` [PATCH v2 09/10] asm-generic: percpu: Add assembly guard Tiwei Bie
2025-09-10 12:12 ` Johannes Berg
2025-08-10 5:51 ` [PATCH v2 10/10] um: Enable SMP support on x86 Tiwei Bie
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=61ae09df-d65b-4c9d-a0c1-7de915246590@t-8ch.de \
--to=linux@weissschuh.net \
--cc=anton.ivanov@cambridgegreys.com \
--cc=arnd@arndb.de \
--cc=benjamin@sipsolutions.net \
--cc=johannes@sipsolutions.net \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-um@lists.infradead.org \
--cc=richard@nod.at \
--cc=tiwei.bie@linux.dev \
--cc=tiwei.btw@antgroup.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).