[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[PATCH 07/10] hurd: Fix x86_64 _hurd_tls_fork
From: |
Sergey Bugaev |
Subject: |
[PATCH 07/10] hurd: Fix x86_64 _hurd_tls_fork |
Date: |
Wed, 17 May 2023 22:14:33 +0300 |
It is illegal to call thread_get_state () on mach_thread_self (), so
this codepath cannot be used as-is to fork the calling thread's TLS.
Fortunately we can use THREAD_SELF (aka %fs:0x0) to find out the value
of our fs_base without calling into the kernel.
Fixes: f6cf701efc61c9ad910372bda14b9a235db310a8
"hurd: Implement TLS for x86_64"
Checked on x86_64-gnu: fork () now works!
Signed-off-by: Sergey Bugaev <bugaevc@gmail.com>
---
It would be nice if GNU Mach allowed
thread_get_state (mach_thread_self (), i386_FSGS_BASE_STATE)
like it already allows thread_set_state for this case. But even if
gnumach adds suppport for this, doing it in userspace without an extra
RPC is still better.
sysdeps/mach/hurd/x86_64/tls.h | 25 +++++++++++++++++++------
1 file changed, 19 insertions(+), 6 deletions(-)
diff --git a/sysdeps/mach/hurd/x86_64/tls.h b/sysdeps/mach/hurd/x86_64/tls.h
index 35dcef44..6487ed35 100644
--- a/sysdeps/mach/hurd/x86_64/tls.h
+++ b/sysdeps/mach/hurd/x86_64/tls.h
@@ -140,12 +140,25 @@ _hurd_tls_fork (thread_t child, thread_t orig,
error_t err;
struct i386_fsgs_base_state state;
mach_msg_type_number_t state_count = i386_FSGS_BASE_STATE_COUNT;
- err = __thread_get_state (orig, i386_FSGS_BASE_STATE,
- (thread_state_t) &state,
- &state_count);
- if (err)
- return err;
- assert (state_count == i386_FSGS_BASE_STATE_COUNT);
+
+ extern thread_t hurd_thread_self (void);
+ if (orig != hurd_thread_self ())
+ {
+ err = __thread_get_state (orig, i386_FSGS_BASE_STATE,
+ (thread_state_t) &state,
+ &state_count);
+ if (err)
+ return err;
+ assert (state_count == i386_FSGS_BASE_STATE_COUNT);
+ }
+ else
+ {
+ /* It is illegal to call thread_get_state () on mach_thread_self ().
+ But we're only interested in the value of fs_base, and since we're
+ this thread, we know it points to our TCB. */
+ state.fs_base = (unsigned long) THREAD_SELF;
+ state.gs_base = 0;
+ }
return __thread_set_state (child, i386_FSGS_BASE_STATE,
(thread_state_t) &state,
--
2.40.1
- Re: [RFC PATCH 06/10] hurd: Make sure to not use tcb->self, (continued)
- Re: [RFC PATCH 06/10] hurd: Make sure to not use tcb->self, Samuel Thibault, 2023/05/17
- Re: [RFC PATCH 06/10] hurd: Make sure to not use tcb->self, Joseph Myers, 2023/05/18
- Re: [RFC PATCH 06/10] hurd: Make sure to not use tcb->self, Sergey Bugaev, 2023/05/18
- Re: [RFC PATCH 06/10] hurd: Make sure to not use tcb->self, Joseph Myers, 2023/05/18
- Re: [RFC PATCH 06/10] hurd: Make sure to not use tcb->self, Samuel Thibault, 2023/05/18
- Re: [RFC PATCH 06/10] hurd: Make sure to not use tcb->self, Sergey Bugaev, 2023/05/19
- Re: [RFC PATCH 06/10] hurd: Make sure to not use tcb->self, Florian Weimer, 2023/05/19
- Re: [RFC PATCH 06/10] hurd: Make sure to not use tcb->self, Joseph Myers, 2023/05/19
- [PATCH] hurd: Fix using interposable hurd_thread_self, Sergey Bugaev, 2023/05/19
- Re: [PATCH] hurd: Fix using interposable hurd_thread_self, Samuel Thibault, 2023/05/19
[PATCH 07/10] hurd: Fix x86_64 _hurd_tls_fork,
Sergey Bugaev <=
[PATCH 08/10] hurd: Fix setting up pthreads, Sergey Bugaev, 2023/05/17
[PATCH 09/10] hurd: Also make it possible to call strlen very early, Sergey Bugaev, 2023/05/17
[RFC PATCH 10/10] hurd: Regenerate errno.h, Sergey Bugaev, 2023/05/17