From: Ian Lance Taylor <iant@google.com>
To: "Björn Schäpers" <gcc@hazardy.de>
Cc: gcc-patches@gcc.gnu.org, gcc@gcc.gnu.org
Subject: Re: [PATCH 4/4] libbacktrace: get debug information for loaded dlls
Date: Thu, 30 Nov 2023 11:53:54 -0800 [thread overview]
Message-ID: <CAKOQZ8xgFYXwvm-=Q6L3r56-cn0Zi27H9P9LxNDrhArLDWPo3A@mail.gmail.com> (raw)
In-Reply-To: <20230120105409.54949-4-gcc@hazardy.de>
On Fri, Jan 20, 2023 at 2:55 AM Björn Schäpers <gcc@hazardy.de> wrote:
>
> From: Björn Schäpers <bjoern@hazardy.de>
>
> Fixes https://github.com/ianlancetaylor/libbacktrace/issues/53, except
> that libraries loaded after the backtrace_initialize are not handled.
> But as far as I can see that's the same for elf.
Thanks, but I don't want a patch that loops using goto statements.
Please rewrite to avoid that. It may be simpler to call a function.
Also starting with a module count of 1000 seems like a lot. Do
typical Windows programs load that many modules?
Ian
> Tested on x86_64-linux and i686-w64-mingw32.
>
> -- >8 --
>
> * pecoff.c (coff_add): New argument for the module handle of the
> file, to get the base address.
> * pecoff.c (backtrace_initialize): Iterate over loaded libraries
> and call coff_add.
>
> Signed-off-by: Björn Schäpers <bjoern@hazardy.de>
> ---
> libbacktrace/pecoff.c | 76 ++++++++++++++++++++++++++++++++++++++++---
> 1 file changed, 72 insertions(+), 4 deletions(-)
>
> diff --git a/libbacktrace/pecoff.c b/libbacktrace/pecoff.c
> index 296f1357b5f..40395109e51 100644
> --- a/libbacktrace/pecoff.c
> +++ b/libbacktrace/pecoff.c
> @@ -49,6 +49,7 @@ POSSIBILITY OF SUCH DAMAGE. */
> #endif
>
> #include <windows.h>
> +#include <psapi.h>
> #endif
>
> /* Coff file header. */
> @@ -592,7 +593,8 @@ coff_syminfo (struct backtrace_state *state, uintptr_t addr,
> static int
> coff_add (struct backtrace_state *state, int descriptor,
> backtrace_error_callback error_callback, void *data,
> - fileline *fileline_fn, int *found_sym, int *found_dwarf)
> + fileline *fileline_fn, int *found_sym, int *found_dwarf,
> + uintptr_t module_handle ATTRIBUTE_UNUSED)
> {
> struct backtrace_view fhdr_view;
> off_t fhdr_off;
> @@ -623,7 +625,6 @@ coff_add (struct backtrace_state *state, int descriptor,
> int is_64;
> uintptr_t image_base;
> uintptr_t base_address = 0;
> - uintptr_t module_handle;
> struct dwarf_sections dwarf_sections;
>
> *found_sym = 0;
> @@ -871,7 +872,6 @@ coff_add (struct backtrace_state *state, int descriptor,
> }
>
> #ifdef HAVE_WINDOWS_H
> - module_handle = (uintptr_t) GetModuleHandleW (NULL);
> base_address = module_handle - image_base;
> #endif
>
> @@ -914,12 +914,80 @@ backtrace_initialize (struct backtrace_state *state,
> int found_sym;
> int found_dwarf;
> fileline coff_fileline_fn;
> + uintptr_t module_handle = 0;
> +
> +#ifdef HAVE_WINDOWS_H
> + DWORD i;
> + DWORD module_count;
> + DWORD bytes_needed_for_modules;
> + HMODULE *modules;
> + char module_name[MAX_PATH];
> + int module_found_sym;
> + fileline module_fileline_fn;
> +
> + module_handle = (uintptr_t) GetModuleHandleW (NULL);
> +#endif
>
> ret = coff_add (state, descriptor, error_callback, data,
> - &coff_fileline_fn, &found_sym, &found_dwarf);
> + &coff_fileline_fn, &found_sym, &found_dwarf, module_handle);
> if (!ret)
> return 0;
>
> +#ifdef HAVE_WINDOWS_H
> + module_count = 1000;
> + alloc_modules:
> + modules = backtrace_alloc (state, module_count * sizeof(HMODULE),
> + error_callback, data);
> + if (modules == NULL)
> + goto skip_modules;
> + if (!EnumProcessModules (GetCurrentProcess (), modules, module_count,
> + &bytes_needed_for_modules))
> + {
> + error_callback(data, "Could not enumerate process modules",
> + (int) GetLastError ());
> + goto free_modules;
> + }
> + if (bytes_needed_for_modules > module_count * sizeof(HMODULE))
> + {
> + backtrace_free (state, modules, module_count * sizeof(HMODULE),
> + error_callback, data);
> + // Add an extra of 2, if some module is loaded in another thread.
> + module_count = bytes_needed_for_modules / sizeof(HMODULE) + 2;
> + modules = NULL;
> + goto alloc_modules;
> + }
> +
> + for (i = 0; i < bytes_needed_for_modules / sizeof(HMODULE); ++i)
> + {
> + if (GetModuleFileNameA (modules[i], module_name, MAX_PATH - 1))
> + {
> + if (strcmp (filename, module_name) == 0)
> + continue;
> +
> + module_handle = (uintptr_t) GetModuleHandleA (module_name);
> + if (module_handle == 0)
> + continue;
> +
> + descriptor = backtrace_open (module_name, error_callback, data, NULL);
> + if (descriptor < 0)
> + continue;
> +
> + coff_add (state, descriptor, error_callback, data,
> + &module_fileline_fn, &module_found_sym, &found_dwarf,
> + module_handle);
> + if (module_found_sym)
> + found_sym = 1;
> + }
> + }
> +
> + free_modules:
> + if (modules)
> + backtrace_free(state, modules, module_count * sizeof(HMODULE),
> + error_callback, data);
> + modules = NULL;
> + skip_modules:
> +#endif
> +
> if (!state->threaded)
> {
> if (found_sym)
> --
> 2.38.1
>
next prev parent reply other threads:[~2023-11-30 19:54 UTC|newest]
Thread overview: 52+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-01-20 10:54 [PATCH 1/4] libbacktrace: change all pc related variables to uintptr_t Björn Schäpers
2023-01-20 10:54 ` [PATCH 2/4] libbacktrace: detect executable path on windows Björn Schäpers
2023-01-23 23:00 ` Ian Lance Taylor
2023-01-24 13:11 ` Eli Zaretskii
2023-01-24 14:35 ` Ian Lance Taylor
2023-01-24 16:52 ` Eli Zaretskii
2023-01-24 17:58 ` Ian Lance Taylor
2023-01-24 18:11 ` Eli Zaretskii
2023-01-24 18:32 ` Ian Lance Taylor
2023-02-05 9:20 ` Björn Schäpers
2023-02-06 0:22 ` Ian Lance Taylor
2023-11-20 19:56 ` Björn Schäpers
2023-11-29 22:05 ` Ian Lance Taylor
2023-01-24 21:00 ` Björn Schäpers
2023-01-20 10:54 ` [PATCH 3/4] libbacktrace: work with aslr " Björn Schäpers
2023-01-20 13:39 ` Eli Zaretskii
2023-01-20 16:46 ` Gabriel Ravier
2023-01-20 19:19 ` Eli Zaretskii
2023-01-20 20:39 ` Gabriel Ravier
2023-01-21 4:05 ` Eli Zaretskii
2023-01-21 9:18 ` LIU Hao
2023-01-21 9:25 ` Eli Zaretskii
2023-01-21 10:47 ` Gabriel Ravier
2023-01-21 11:42 ` Eli Zaretskii
2023-11-20 19:57 ` Björn Schäpers
2023-11-20 20:07 ` Eli Zaretskii
2023-11-21 19:35 ` Björn Schäpers
2023-11-22 1:13 ` LIU Hao
2023-11-30 19:25 ` Ian Lance Taylor
2023-01-20 10:54 ` [PATCH 4/4] libbacktrace: get debug information for loaded dlls Björn Schäpers
2023-11-30 19:53 ` Ian Lance Taylor [this message]
2023-11-30 20:16 ` Eli Zaretskii
2024-01-02 23:12 ` Björn Schäpers
2024-01-04 22:33 ` [PATCH 5/4] libbacktrace: improve getting " Björn Schäpers
2024-01-06 22:15 ` [PATCH 6/4] libbacktrace: Add loaded dlls after initialize Björn Schäpers
2024-01-07 6:50 ` Eli Zaretskii
[not found] ` <4cb3a2a5-c0b3-40c8-b460-f21d65a9aea2@hazardy.de>
2024-01-07 14:46 ` Eli Zaretskii
2024-01-07 16:07 ` Björn Schäpers
2024-01-07 17:03 ` Eli Zaretskii
2024-01-09 20:02 ` Björn Schäpers
2024-01-10 12:34 ` Eli Zaretskii
2024-03-15 20:41 ` Björn Schäpers
2024-01-23 22:37 ` [PATCH 5/4] libbacktrace: improve getting debug information for loaded dlls Ian Lance Taylor
2024-01-25 19:53 ` Björn Schäpers
2024-01-25 22:04 ` Ian Lance Taylor
2024-03-15 20:40 ` Björn Schäpers
2024-04-25 20:14 ` Björn Schäpers
2024-04-28 18:16 ` Ian Lance Taylor
2024-05-02 19:23 ` Björn Schäpers
2024-01-23 21:24 ` [PATCH 4/4] libbacktrace: get " Björn Schäpers
2023-01-20 22:25 ` [PATCH 1/4] libbacktrace: change all pc related variables to uintptr_t Ian Lance Taylor
2023-01-23 20:17 ` Björn Schäpers
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='CAKOQZ8xgFYXwvm-=Q6L3r56-cn0Zi27H9P9LxNDrhArLDWPo3A@mail.gmail.com' \
--to=iant@google.com \
--cc=gcc-patches@gcc.gnu.org \
--cc=gcc@gcc.gnu.org \
--cc=gcc@hazardy.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;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).