Re: [RFC PATCH glibc 1/4] glibc: Perform rseq(2) registration at nptl init and thread creation (v4)

----- On Dec 4, 2018, at 2:44 PM, Florian Weimer fweimer@redhat.com wrote:

* Mathieu Desnoyers:

quoted

diff --git a/sysdeps/unix/sysv/linux/riscv/rv64/libpthread.abilist

b/sysdeps/unix/sysv/linux/riscv/rv64/libpthread.abilist
index c370fda73d..9da78d59d2 100644

--- a/sysdeps/unix/sysv/linux/riscv/rv64/libpthread.abilist
+++ b/sysdeps/unix/sysv/linux/riscv/rv64/libpthread.abilist

@@ -235,3 +235,4 @@ GLIBC_2.28 tss_create F
 GLIBC_2.28 tss_delete F
 GLIBC_2.28 tss_get F
 GLIBC_2.28 tss_set F
+GLIBC_2.29 __rseq_refcount T 0x4

This part looks buggy.

It's unclear based on this patch whether you actually want to get rid of
the __rseq_refcount symbol.

I want to keep the __rseq_refcount symbol so out-of-libc users can
register rseq if they are linked against a pre-2.29 libc.

Why does it look buggy ?

Thanks,

Mathieu


-- 
Mathieu Desnoyers
EfficiOS Inc.
http://www.efficios.com