From: Elena Zannoni <ezannoni@redhat.com>
To: Elena Zannoni <ezannoni@redhat.com>
Cc: gdb-patches@sources.redhat.com
Subject: Re: [PATCH/RFC] TLS support part 1
Date: Mon, 21 Oct 2002 13:39:00 -0000 [thread overview]
Message-ID: <15796.25948.439753.439692@localhost.redhat.com> (raw)
In-Reply-To: <15770.25139.262675.365054@localhost.redhat.com>
Elena Zannoni writes:
>
> This part of the tls changes deal with the symbol handling.
> I.e. recognizes the debug information, and sets up variables with the
> appropriate address information. Then it requests the address to be
> computed by the target/thread code.
>
> BTW: the debug info looks like this (in a simple case):
>
> <1><1e2c>: Abbrev Number: 32 (DW_TAG_variable)
> DW_AT_name : (indirect string, offset: 0xc0d): a_tls
> DW_AT_decl_file : 1
> DW_AT_decl_line : 4
> DW_AT_type : <c75>
> DW_AT_external : 1
> DW_AT_location : 6 byte block: 3 0 0 0 0 e0 (DW_OP_addr: 0; DW_OP_GNU_push_tls_address; )
> <1><1e3f>: Abbrev Number: 32 (DW_TAG_variable)
> DW_AT_name : (indirect string, offset: 0xb42): b_tls
> DW_AT_decl_file : 1
> DW_AT_decl_line : 5
> DW_AT_type : <c75>
> DW_AT_external : 1
> DW_AT_location : 6 byte block: 3 4 0 0 0 e0 (DW_OP_addr: 4; DW_OP_GNU_push_tls_address; )
> <1><1e52>: Abbrev Number: 32 (DW_TAG_variable)
> DW_AT_name : (indirect string, offset: 0xaa8): c_tls
> DW_AT_decl_file : 1
> DW_AT_decl_line : 6
> DW_AT_type : <c75>
> DW_AT_external : 1
> DW_AT_location : 6 byte block: 3 8 0 0 0 e0 (DW_OP_addr: 8; DW_OP_GNU_push_tls_address; )
>
> I.e. a_tls, b_tls, c_tls are thread local storage variables at offset
> 0, 4, 8 respectively within the thread local storage.
>
> The next patch will deal with the actual computation of the address.
>
> Elena
>
> 2002-10-01 Jim Blandy <jimb@redhat.com>
> Elena Zannoni <ezannoni@redhat.com>
>
> * symtab.h (address_class): Re-add LOC_THREAD_LOCAL_STATIC
> for thread local storage locations.
> (struct symbol): Add objfile field.
> (SYMBOL_OBJFILE): Define.
> * dwarf2read.c (is_thread_local): New static variable.
> (new_symbol): If variable is in thread local fill in address class
> and objfile appropriately.
> (decode_locdesc): Recognize and handle DW_OP_GNU_push_tls_address
> stack operation.
> * printcmd.c (address_info): Print the information for thread
> local storage variable.
> * findvar.c (read_var_value): In case of thread local variable,
> defer to the target vector code to compute address.
[forgot to send mail].
I have committed this (after updating the comment per Jim's request).
Elena
prev parent reply other threads:[~2002-10-21 20:39 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2002-10-01 20:06 Elena Zannoni
2002-10-02 18:37 ` Jim Blandy
2002-10-02 18:49 ` Elena Zannoni
2002-10-21 13:39 ` Elena Zannoni [this message]
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=15796.25948.439753.439692@localhost.redhat.com \
--to=ezannoni@redhat.com \
--cc=gdb-patches@sources.redhat.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