mail archive of the barebox mailing list
 help / color / mirror / Atom feed
From: Ahmad Fatoum <a.fatoum@pengutronix.de>
To: barebox@lists.infradead.org
Cc: Peter Mamonov <pmamonov@gmail.com>,
	Peter Korsgaard <peter@korsgaard.com>,
	Sascha Hauer <sha@pengutronix.de>
Subject: [PATCH v3 01/16] common: introduce HAS_ARCH_SJLJ
Date: Wed, 10 Mar 2021 09:47:45 +0100	[thread overview]
Message-ID: <20210310084800.3584-2-a.fatoum@pengutronix.de> (raw)
In-Reply-To: <20210310084800.3584-1-a.fatoum@pengutronix.de>

From: Ahmad Fatoum <ahmad@a3f.at>

setjmp stores the registers comprising the execution context into a
jmp_buf and longjmp switches to that context and continues execution just
after the setjmp that allocated that jmp_buf.

setjmp/longjmp can have multiple uses:

  - Handle return from firmware that clobbers registers. This is why we
    already have a setjmp/longjmp ARM

  - Do exception handling[1]

  - Port over scripting language VMs that use sjlj for context switching

  - Implement stackless coroutines in C.

These coroutines would be stackless, because jumping to a setjmp down the
call stack means that the code there will clobber the stack below it.
On resuming the coroutine, it will run with a stack changed in the
interim leading to undefined behavior.

There are ways around that without resorting to custom Assembly:

  - Allocate a buffer on the scheduler's stack, so coroutine can
    grow into them
     -> Problem: exploits Undefined behavior

  - Yield first time on scheduler stack, then patch jmp_buf to point at
    another stack
     -> Problem: Code switching stacks should not itself use the stack

  - Construct a new jmp_buf with user-supplied return address and stack top:
    int initjmp(jmp_buf jmp, void __noreturn (*func)(void), void *stack_top);

The latter could also be reused as a portable way of setting up a stack
in early init code.

We are going to implement green threads in later commits, so add a
symbol that can be selected to indicate architecture support for setjmp,
longjmp and the new initjmp.

[1]: https://www.spinics.net/lists/u-boot-v2/msg41322.html

Signed-off-by: Ahmad Fatoum <ahmad@a3f.at>
---
 common/Kconfig | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/common/Kconfig b/common/Kconfig
index edadcc9f4979..c0ff57bcdba4 100644
--- a/common/Kconfig
+++ b/common/Kconfig
@@ -28,6 +28,11 @@ config HAS_DMA
 	  Drivers that depend on a DMA implementation can depend on this
 	  config, so that you don't get a compilation error.
 
+config HAS_ARCH_SJLJ
+	bool
+	help
+	  Architecture has support implemented for setjmp()/longjmp()/initjmp()
+
 config GENERIC_GPIO
 	bool
 
-- 
2.29.2


_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox


  reply	other threads:[~2021-03-10  8:49 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-03-10  8:47 [PATCH v3 00/16] common: introduce bthreads, co-operative Ahmad Fatoum
2021-03-10  8:47 ` Ahmad Fatoum [this message]
2021-03-10  8:47 ` [PATCH v3 02/16] ARM: asm: setjmp: annotate setjmp/longjmp for GCC Ahmad Fatoum
2021-03-10  8:47 ` [PATCH v3 03/16] ARM: asm: setjmp: implement coroutine dependency initjmp() Ahmad Fatoum
2021-03-10  8:47 ` [PATCH v3 04/16] sandbox: asm: implement setjmp/longjmp/initjmp Ahmad Fatoum
2021-03-10  8:47 ` [PATCH v3 05/16] riscv: Add asm/asm.h Ahmad Fatoum
2021-03-10  8:47 ` [PATCH v3 06/16] riscv: Add asm/linkage.h Ahmad Fatoum
2021-03-10  8:47 ` [PATCH v3 07/16] riscv: Implement setjmp/longjmp/initjmp for RV32I Ahmad Fatoum
2021-03-10  8:47 ` [PATCH v3 08/16] mips: Add linkage.h Ahmad Fatoum
2021-03-10  8:47 ` [PATCH v3 09/16] mips: Implement setjmp/longjmp/initjmp for 32BIT Ahmad Fatoum
2021-03-10  8:47 ` [PATCH v3 10/16] powerpc: Implement initjmp/setjmp/longjmp Ahmad Fatoum
2021-03-10  8:47 ` [PATCH v3 11/16] x86: implement setjmp/longjmp/initjmp Ahmad Fatoum
2021-03-10  8:47 ` [PATCH v3 12/16] common: poller: replace explicit calls to poller_call() with resched() Ahmad Fatoum
2021-03-10  8:47 ` [PATCH v3 13/16] console: unconditionally run resched() in ctrlc() Ahmad Fatoum
2021-03-10  8:47 ` [PATCH v3 14/16] common: introduce bthreads, co-operative barebox threads Ahmad Fatoum
2021-03-10  8:47 ` [PATCH v3 15/16] commands: add new bthread test command Ahmad Fatoum
2021-03-10  8:48 ` [PATCH v3 16/16] RFC: led: migrate from poller to bthread Ahmad Fatoum
2021-03-15  8:34 ` [PATCH v3 00/16] common: introduce bthreads, co-operative Sascha Hauer

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=20210310084800.3584-2-a.fatoum@pengutronix.de \
    --to=a.fatoum@pengutronix.de \
    --cc=barebox@lists.infradead.org \
    --cc=peter@korsgaard.com \
    --cc=pmamonov@gmail.com \
    --cc=sha@pengutronix.de \
    /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