Skip to contentSkip to conversation

Adding google-breakpad to buildroot

Buildroot Pascal Hürst 3 participants 6 messages Apr 23, 2014 – Apr 24, 2014

lore
Share & tools
lore
#0 lore raw

Hi everyone,

we are planing to add google-breakpad to buildroot.

From the project description at
http://code.google.com/p/google-breakpad/:

[...] "Breakpad is a library and tool suite that allows you to distribute an application to users with compiler-provided debugging information removed, record crashes in compact "minidump" files, send them back to your server, and produce C and C++ stack traces from these minidumps." [...]

Adding a package to buildroot is easy, but in this case we will have to find a way to extract all symbols from the target binaries, before they get stripped. My idea was to add a new target to the Makefile, just before "target-finalize:" and check against an option in the config like:

ifeq ($(BR2_ENABLE_BREAKPAD),y)

        TARGETS+=target-generate-breakpad-symbols

endif

and then:

target-generate-breakpad-symbols:
extract symbols and deploy result to output/images

Is this basically the way to go, or is there a better way to achieve this?

regards,
pascal
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 198 bytes
Desc: This is a digitally signed message part
URL: <http://lists.busybox.net/pipermail/buildroot/attachments/20140423/eebac11d/attachment.asc>

#1 lore raw
Quoted 23 lines · Pascal H?rst

Hi everyone,

we are planing to add google-breakpad to buildroot.

From the project description at
http://code.google.com/p/google-breakpad/:

[...] "Breakpad is a library and tool suite that allows you to distribute an application to users with compiler-provided debugging information removed, record crashes in compact "minidump" files, send them back to your server, and produce C and C++ stack traces from these minidumps." [...]

Adding a package to buildroot is easy, but in this case we will have to find a way to extract all symbols from the target binaries, before they get stripped. My idea was to add a new target to the Makefile, just before "target-finalize:" and check against an option in the config like:

ifeq ($(BR2_ENABLE_BREAKPAD),y)

        TARGETS+=target-generate-breakpad-symbols

endif

Quoted 7 lines

and then:

target-generate-breakpad-symbols:
extract symbols and deploy result to output/images

Is this basically the way to go, or is there a better way to achieve this?

It is easier to add it to the target-finalize target. Then you can be
sure that the ordering is correct (target-finalize already has all the needed dependencies). You can also easily use conditions there (it's not inside a define).

However, I think it will be more appropriate to implement breakpad as an
additional strip alternative. You probably don't want to combine strip with breakpad... Currently we have none, strip or sstrip, it should be relatively easy to add breakpad as an additional option. Note, however, that the strip code is very old, it doesn't satisfy our coding style so we'll probably want you to do some cleanup first. Also, it is currently untested in the autobuilders AFAIK.

Regards,
Arnout

Regards,
Arnout

Signature · 14 lines
Arnout Vandecappelle                          arnout at mind be
Senior Embedded Software Architect            +32-16-286500
Essensium/Mind                                http://www.mind.be
G.Geenslaan 9, 3001 Leuven, Belgium           BE 872 984 063 RPR Leuven
LinkedIn profile: http://www.linkedin.com/in/arnoutvandecappelle
GPG fingerprint:  7CB5 E4CC 6C2E EFD4 6E3D A754 F963 ECAB 2450 2F1F

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 880 bytes
Desc: OpenPGP digital signature
URL: <http://lists.busybox.net/pipermail/buildroot/attachments/20140423/bbd825f8/attachment.asc>
#2 lore raw

Hi Pascal, Arnout, all,

Quoted 14 lines · Arnout Vandecappelle
Quoted 13 lines · Pascal H?rst

Hi everyone,

we are planing to add google-breakpad to buildroot.

From the project description at
http://code.google.com/p/google-breakpad/:

[...] "Breakpad is a library and tool suite that allows you to distribute an application to users with compiler-provided debugging information removed, record crashes in compact "minidump" files, send them back to your server, and produce C and C++ stack traces from these minidumps." [...]

Quoted 3 lines
Quoted 3 lines

Adding a package to buildroot is easy, but in this case we will have to find a way to extract all symbols from the target binaries, before they get stripped.

Here, you mean adding the host-breakpad package, right?

IIRC, the handler in integrated at the source code level in the projects. Do you plan to provide something for the target?

Quoted 16 lines
Quoted 8 lines

My idea was to add a new target to the Makefile, just before "target-finalize:" and check against an option in the config like:

ifeq ($(BR2_ENABLE_BREAKPAD),y)

        TARGETS+=target-generate-breakpad-symbols

endif

Quoted 7 lines

and then:

target-generate-breakpad-symbols:
extract symbols and deploy result to output/images

Is this basically the way to go, or is there a better way to achieve this?

Quoted 8 lines

It is easier to add it to the target-finalize target. Then you can be
sure that the ordering is correct (target-finalize already has all the needed dependencies). You can also easily use conditions there (it's not inside a define).

However, I think it will be more appropriate to implement breakpad as an
additional strip alternative. You probably don't want to combine strip with breakpad...

I think breakpad is more like a pre-strip hook, that one can enable or not. But in the end, the target image will be stripped most of the time (at least, this is the way I would use such a feature).

Quoted 6 lines

Currently we have none, strip or sstrip, it should be relatively easy to add breakpad as an additional option. Note, however, that the strip code is very old, it doesn't satisfy our coding style so we'll probably want you to do some cleanup first. Also, it is currently untested in the autobuilders AFAIK.

Regards,

Signature · 1 line
Samuel
#3 lore raw
Quoted 17 lines · Samuel Martin

Hi Pascal, Arnout, all,

Quoted 14 lines · Arnout Vandecappelle
Quoted 13 lines · Pascal H?rst

Hi everyone,

we are planing to add google-breakpad to buildroot.

From the project description at
http://code.google.com/p/google-breakpad/:

[...] "Breakpad is a library and tool suite that allows you to distribute an application to users with compiler-provided debugging information removed, record crashes in compact "minidump" files, send them back to your server, and produce C and C++ stack traces from these minidumps." [...]

Quoted 5 lines
Quoted 3 lines
Quoted 3 lines

Adding a package to buildroot is easy, but in this case we will have to find a way to extract all symbols from the target binaries, before they get stripped.

Here, you mean adding the host-breakpad package, right?

I acutally mean both sides. Right now it (*.mk) looks something like that:

GOOGLE_BREAKPAD_VERSION = 1276
GOOGLE_BREAKPAD_SITE = http://google-breakpad.googlecode.com/svn/trunk
GOOGLE_BREAKPAD_SITE_METHOD = svn

GOOGLE_BREAKPAD_CONF_OPT = --disable-processor --disable-tools $(eval $(host-autotools-package))
$(eval $(autotools-package))

Quoted 2 lines

IIRC, the handler in integrated at the source code level in the projects. Do you plan to provide something for the target?

exactly, host and target are needed

Quoted 21 lines
Quoted 21 lines
Quoted 8 lines

My idea was to add a new target to the Makefile, just before "target-finalize:" and check against an option in the config like:

ifeq ($(BR2_ENABLE_BREAKPAD),y)

        TARGETS+=target-generate-breakpad-symbols

endif

Quoted 7 lines

and then:

target-generate-breakpad-symbols:
extract symbols and deploy result to output/images

Is this basically the way to go, or is there a better way to achieve this?

It is easier to add it to the target-finalize target. Then you can be
sure that the ordering is correct (target-finalize already has all the needed dependencies). You can also easily use conditions there (it's not inside a define).

Quoted 7 lines
Quoted 3 lines

However, I think it will be more appropriate to implement breakpad as an
additional strip alternative. You probably don't want to combine strip with breakpad...

I think breakpad is more like a pre-strip hook, that one can enable or not. But in the end, the target image will be stripped most of the time (at least, this is the way I would use such a feature).

I agree, since the extraction of the symbols should always happen, when breakpad is used, no mater what strip alternative is selected

Quoted 10 lines
Quoted 6 lines

Currently we have none, strip or sstrip, it should be relatively easy to add breakpad as an additional option. Note, however, that the strip code is very old, it doesn't satisfy our coding style so we'll probably want you to do some cleanup first. Also, it is currently untested in the autobuilders AFAIK.

Regards,

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 198 bytes
Desc: This is a digitally signed message part
URL: <http://lists.busybox.net/pipermail/buildroot/attachments/20140423/a29df25c/attachment.asc>

#4 lore raw
Quoted 5 lines · Pascal H?rst
Quoted 4 lines · Samuel Martin

Hi Pascal, Arnout, all,

Quoted 1 line · Arnout Vandecappelle

On 23/04/14 16:21, Pascal H?rst wrote:

[snip]

Quoted 9 lines
Quoted 9 lines
Quoted 8 lines

It is easier to add it to the target-finalize target. Then you can be
sure that the ordering is correct (target-finalize already has all the needed dependencies). You can also easily use conditions there (it's not inside a define).

However, I think it will be more appropriate to implement breakpad as an
additional strip alternative. You probably don't want to combine strip with breakpad...

Quoted 6 lines
Quoted 3 lines

I think breakpad is more like a pre-strip hook, that one can enable or not. But in the end, the target image will be stripped most of the time (at least, this is the way I would use such a feature).

I agree, since the extraction of the symbols should always happen, when breakpad is used, no mater what strip alternative is selected

OK, I thought breakpad did stripping already.

In that case, I think it's best to put it in target-finalize just before
strip.

Regards,
Arnout

Signature · 14 lines
Arnout Vandecappelle                          arnout at mind be
Senior Embedded Software Architect            +32-16-286500
Essensium/Mind                                http://www.mind.be
G.Geenslaan 9, 3001 Leuven, Belgium           BE 872 984 063 RPR Leuven
LinkedIn profile: http://www.linkedin.com/in/arnoutvandecappelle
GPG fingerprint:  7CB5 E4CC 6C2E EFD4 6E3D A754 F963 ECAB 2450 2F1F

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 880 bytes
Desc: OpenPGP digital signature
URL: <http://lists.busybox.net/pipermail/buildroot/attachments/20140423/84180dd6/attachment.asc>
#5 lore raw
Quoted 20 lines · Arnout Vandecappelle
Quoted 5 lines · Pascal H?rst
Quoted 4 lines · Samuel Martin

Hi Pascal, Arnout, all,

Quoted 1 line · Arnout Vandecappelle

On 23/04/14 16:21, Pascal H?rst wrote:

[snip]

Quoted 13 lines
Quoted 12 lines
Quoted 8 lines

It is easier to add it to the target-finalize target. Then you can be
sure that the ordering is correct (target-finalize already has all the needed dependencies). You can also easily use conditions there (it's not inside a define).

However, I think it will be more appropriate to implement breakpad as an
additional strip alternative. You probably don't want to combine strip with breakpad...

I think breakpad is more like a pre-strip hook, that one can enable or not. But in the end, the target image will be stripped most of the time (at least, this is the way I would use such a feature).

Quoted 7 lines
Quoted 2 lines

I agree, since the extraction of the symbols should always happen, when breakpad is used, no mater what strip alternative is selected

OK, I thought breakpad did stripping already.

In that case, I think it's best to put it in target-finalize just before
strip.

Ok, I'll do it that way, then.

Thanks!

Quoted 2 lines

Regards,
Arnout

Regards
Pascal

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 198 bytes
Desc: This is a digitally signed message part
URL: <http://lists.busybox.net/pipermail/buildroot/attachments/20140424/990dae1f/attachment.asc>