[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Wed, 10 Jun 2015 23:02:35 -0500
From: Rob Landley <rob@...dley.net>
To: Rich Felker <dalias@...c.org>, musl@...ts.openwall.com,
"D. Jeff Dionne" <Jeff@...inux.org>,
ysato@...rs.sourceforge.jp, shumpei.kawasaki@...wc.com
Subject: Re: Moving forward with sh2/nommu
On 06/09/2015 10:30 PM, Rich Felker wrote:
> On Mon, Jun 01, 2015 at 11:11:07AM -0400, Rich Felker wrote:
>> [resent to musl list]
>>
>> Here's a summary of the issues we need to work through to get a modern
>> SH2/nommu-targetted musl/toolchain out of the proof-of-concept stage
>> and to the point where it's something people can use roughly 'out of
>> the box':
>>
>> Kernel issues:
>>
>> 1. Kernel should support loading plain ELF directly, unmodified. Right
>
> I have a patch to do this, not polished but it works. It also...
>
>> 2. Kernel insists on having a stack size set in the PT_GNU_STACK
>> program header; if it's 0 (the default ld produces) then execve
>> fails. It should just provide a default, probably 128k (equal to
>> MMU-ful Linux).
>
> ...uses a default stack size of 128k if the header is 0, and...
That's big enough to give a false sense of security without being big
enough to actually work reliably on software that isn't thinking about it.
The kernel stack size is apparently 4k, but they've got more than one
for different contexts:
https://www.kernel.org/doc/Documentation/x86/x86_64/kernel-stacks
I think I agree to jeff here: the historical default stack size for
nommu packaging was 8k. That's enough for hello world to work, but in
this context stack is a resource we need to allocate if we want
nontrivial amounts of it.
>> 3. Kernel uses the stack for brk too, growing brk from the opposite
>> end. This is horribly buggy/dangerous. Just dummying out brk to
>
> ...set the brk size to zero so that brk will always fail.
Yay always failing. The brk approach does not make sense in this context.
>> 4. Syscall trap numbers differ on SH2 vs SH3/4. Presumably the reason
>> is that these two SH2A hardware traps overlap with the syscall
>> range used by SH3/4 ABI:
>
> I haven't patched this yet. I'd like to use 31 (0x1f) as the new
> universal SH syscall trap number, instead of 22. More details on the
> reasons later.
I've cc'd Yoshinori Sato (who did most of the historical sh2 work) and
Shumpei Kawasaki (the original superh architect). They'll probably have
an opinion on your "more reasons" for changing sh2 system call numbers
to match sh4.
>> musl issues:
>>
>> 1. We need runtime detection for the right trap number to use for
>> syscalls. Right now I've got the trap numbers hard-coded for SH2 in
>> my local tree.
>
> I've written the runtime detection, but I'd rather not have to use it.
> I managed to avoid inlining a big conditional at each syscall, but
> there are still multiple ugly issues:
This is only an issue if you want sh4 to be able to run sh2 binaries.
(Can m68k run coldfire binaries? Last I checked there wasn't any
blackfin-with-mmu variant.)
Rebuilding your source for a different output format really isn't a big
deal in open source. I know you wanted musl binary compatible with glibc
too, but I don't see the advantage.
> - The cancellable syscall code has two different instruction addresses
> where the actual syscall could take place (depending on which trap
> instruction is used), whereas the current musl C code assumes that
> there's exactly one instruction pointer range where cancellation may
> be acted upon. I have code to allow a second range but it's ugly.
>
> - The sigreturn (sa_restorer) functions need to use runtime selection,
> which precludes putting them in the canonical form they usually
> have. At some point in the past, not using the exact instruction
> sequences expected for these functions confused gdb, but I suspect
> that's a non-issue on modern versions. Alternatively we could have
> sigreturn.c pass different versions of the sigreturn functions to
> the kernel depending on which syscall ABI is in use at runtime, but
> this requires bolting on new arch-specific logic to something that's
> normally not arch-specific, so I prefer just having the conditionals
> in the sigreturn functions.
>
> - The syscall asm requires an extra register clobber to be able to do
> the runtime trap number switching.
>
> An alternate design I considered was simply patching the trap number
> in the .text for SH2, since there's no memory protection, but of
> course this would preclude execute-in-place from ROM, etc. so I don't
> think it's a proper solution. It is much less invasive, though.
We should finagle together an XIP test setup. There were two different
people doing it at ELC we could ask questions of.
>> 2. We need additional runtime detection options for atomics: interrupt
>> masking for plain SH2, and the new CAS instruction for SH2J.
>
> This is the one thing I haven't done, so currently the atomic macros
> are using GUSA which is non-atomic and unsafe on SH2 (if an interrupt
> happens with invalid stack pointer, memory will be corrupted).
And doesn't work with SMP at all.
> This
> could be part of the random crashing I've been experiencing (although
> I reproduced it without musl) so I'll try to add them next.
I'm going to try to post kernel patches to the list later today, and
separately email you the horrible ethernet drivers that aren't going
upstream because horrible. (We need to clean up the ethernet VHDL too,
it has timing issues that randomly work or don't work because layout
butterfly effects. Known problem, on the todo list, not my area...)
>> 3. We need sh/vfork.s since the default vfork.c just uses fork, which
>> won't work. I have a version locally but it doesn't make sense to
>> commit without runtime trap number selection.
>
> Done and updated to use runtime selection in the (ugly) patch.
If they ask for vfork() they should get vfork()...?
>> 4. As long as we're using the FDPIC ELF header flag to get
>> binfmt_elf_fdpic.c to load binaries, the startup code needs to call
>> the personality() syscall to switch back. I have a local hack for
>> doing this in rcrt1.o which is probably not worth upstreaming if we
>> can just make the kernel do it right.
>
> No longer needed because of the kernel patch to load normal ELF.
Send me the patch and I'll add it to my stack to go upstream.
>> 5. The brk workaround I'm doing now can't be upstreamed without a
>> reliable runtime way to distinguish nommu. To put it in malloc.c
>> this would have to be a cross-arch solution. What might make more
>> sense is putting it in syscall_arch.h for sh, where we already
>> have to check for SH2 to determine the right trap number; the
>> inline syscall code can just do if (nr==SYS_brk&&IS_SH2) return 0;
>
> Commit 276904c2f6bde3a31a24ebfa201482601d18b4f9 in musl solves this in
> a general manner, even though it's no longer needed with my kernel
> patch applied.
No longer needed on sh2 maybe, but there are a half-dozen other nommu
targets of interest...
> One more musl-side issue I neglected to mention is the __unmapself.s
> can't work on SH2 because the SH2 trap/interrupt mechanism requires
> the userspace stack pointer to be valid at all times. This is now
> solved upstream in commit c30cbcb0a646b1f13a22c645616dce624465b883,
> but activating it for SH2 requires removing
> src/thread/sh/__unmapself.s so the generic C file gets used.
>
> The attached patch covers everything described above that's not
> already upstream, and is sufficient to build musl for sh2 with
> musl-cross targeting "sheb-linux-musl". I used gcc 4.7.3 because later
> versions break the kernel. The attached config.mak for musl shows the
> configure options I used. The attached sheb.specs file is how I got
> gcc to do always-PIE without breaking the kernel.
For the newly cc'd the relevant web archive entry is:
http://www.openwall.com/lists/musl/2015/06/10/1
Rob
Powered by blists - more mailing lists
Confused about mailing lists and their use?
Read about mailing lists on Wikipedia
and check out these
guidelines on proper formatting of your messages.
