Taking into account BR2_PREFER_STATIC_LIB rename
PATCH 1/2 Taking into account BR2_PREFER_STATIC_LIB rename
In Buildroot, the BR2_PREFER_STATIC_LIB option was renamed to BR2_STATIC_LIBS. This commit takes this into account when generating the autobuild configurations.
--- scripts/autobuild-run | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
scripts/autobuild-run +1 -1
@@ -348,7 +348,7 @@ def gen_config(instance, log, sysinfo): elif randint(0, 20) == 0: configlines.append("BR2_ROOTFS_DEVICE_CREATION_DYNAMIC_EUDEV=y\n") if config["libc"] != "glibc" and randint(0, 20) == 0: - configlines.append("BR2_PREFER_STATIC_LIB=y\n") + configlines.append("BR2_STATIC_LIBS=y\n") # Write out the configuration file with open(os.path.join(outputdir, ".config"), "w+") as configf:
Signature · 1 line
2.1.0
PATCH 1/2 Taking into account BR2_PREFER_STATIC_LIB rename
Hello,
scripts/autobuild-run def gen_config(instance, log, sysinfo): · line 351
Quoted 3 lines · Thomas Petazzoni
if config["libc"] != "glibc" and randint(0, 20) == 0:scripts/autobuild-run +1
-1- configlines.append("BR2_PREFER_STATIC_LIB=y\n") + configlines.append("BR2_STATIC_LIBS=y\n")
Instead, should I also sometimes test BR2_SHARED_STATIC_LIBS ? Like 5% of times static only, 5% of times static+shared, and 90% of the time shared only (the default) ?
Best regards,
Thomas
Signature · 3 lines
Thomas Petazzoni, CTO, Free Electrons Embedded Linux, Kernel and Android engineering http://free-electrons.com
PATCH 1/2 Taking into account BR2_PREFER_STATIC_LIB rename
Thomas, All,
On 2014-12-21 22:03 +0100, Thomas Petazzoni spake thusly:
Quoted 8 lines · Thomas Petazzoni
scripts/autobuild-run def gen_config(instance, log, sysinfo): · line 351
Quoted 3 lines
if config["libc"] != "glibc" and randint(0, 20) == 0:scripts/autobuild-run +1
-1- configlines.append("BR2_PREFER_STATIC_LIB=y\n") + configlines.append("BR2_STATIC_LIBS=y\n")Instead, should I also sometimes test
BR2_SHARED_STATIC_LIBS? Like 5% of times static only, 5% of times static+shared, and 90% of the time shared only (the default) ?
Yes.
If there are some packages that do not like to be build both static and shared at the same time, we want to catch those.
Regards,
Yann E. MORIN.
Signature · 6 lines
.-----------------.--------------------.------------------.--------------------. | Yann E. MORIN | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: | | +33 662 376 056 | Software Designer | \ / CAMPAIGN | ___ | | +33 223 225 172 `------------.-------: X AGAINST | \e/ There is no | | http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL | v conspiracy. | '------------------------------^-------^------------------^--------------------'
PATCH 1/2 Taking into account BR2_PREFER_STATIC_LIB rename
Dear Yann E. MORIN,
Quoted 2 lines · Yann E. MORIN
If there are some packages that do not like to be build both static and shared at the same time, we want to catch those.
I've already pushed my autobuild-run patches, but we can do a followup patch to implement this.
Thanks,
Thomas
Signature · 3 lines
Thomas Petazzoni, CTO, Free Electrons Embedded Linux, Kernel and Android engineering http://free-electrons.com
PATCH 2/2 Enable paranoid path checking
In reply to Thomas PetazzoniNow that Buildroot has support for paranoid path checking (check for clearly invalid header and library directories used when cross-compiling), let's enable this in the autobuilders.
Without this, the paranoid path checking only emits warnings. With this patch enabled, whenever an invalid header or library path is used, the build will stop and fail, which will allow us to clearly notice such cases.
---
Do we want to apply this right now? We already have a lot of failures caused by the static/shared change. Do we want to have more?
--- scripts/autobuild-run | 1 + 1 file changed, 1 insertion(+)
scripts/autobuild-run +1 -0
@@ -341,6 +341,7 @@ def gen_config(instance, log, sysinfo): # Amend the configuration with a few things. configlines.append("BR2_PACKAGE_BUSYBOX_SHOW_OTHERS=y\n") configlines.append("# BR2_TARGET_ROOTFS_TAR is not set\n") + configlines.append("BR2_COMPILER_PARANOID_UNSAFE_PATH=y\n") if randint(0, 20) == 0: configlines.append("BR2_ENABLE_DEBUG=y\n") if randint(0, 30) == 0:
Signature · 1 line
2.1.0
PATCH 2/2 Enable paranoid path checking
Thomas Petazzoni <thomas.petazzoni-wi1+55ScJUtKEb57/3fJTNBPR1lH4CV8 @public.gmane.org> wrote in news:1418852858-818-2-git-send-email- thomas.petazzoni at free-electrons.com:
Quoted 2 lines
Do we want to apply this right now? We already have a lot of failures caused by the static/shared change. Do we want to have more?
[...]
Quoted 1 line
+
configlines.append("BR2_COMPILER_PARANOID_UNSAFE_PATH=y\n")
Hi,
I vote for yes.
Regards, Bernd
PATCH 2/2 Enable paranoid path checking
In reply to Thomas PetazzoniThomas, All,
On 2014-12-17 22:47 +0100, Thomas Petazzoni spake thusly:
Quoted 9 lines
Now that Buildroot has support for paranoid path checking (check for clearly invalid header and library directories used when cross-compiling), let's enable this in the autobuilders.
Without this, the paranoid path checking only emits warnings. With this patch enabled, whenever an invalid header or library path is used, the build will stop and fail, which will allow us to clearly notice such cases.
Quoted 4 lines
Signed-off-by Thomas Petazzoni <[email protected]>---
Do we want to apply this right now? We already have a lot of failures caused by the static/shared change. Do we want to have more?
Well, either we set it right now, so we have time to investigate before the 2015.02 release, or we post-pone that for after.
We're almost one mont in the release cycle, with a bit more than another month before rc1, so it might be already too late for the switch, but it is still doable.
So:
Regards,
Yann E. MORIN.
Signature · 6 lines
.-----------------.--------------------.------------------.--------------------. | Yann E. MORIN | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: | | +33 662 376 056 | Software Designer | \ / CAMPAIGN | ___ | | +33 223 225 172 `------------.-------: X AGAINST | \e/ There is no | | http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL | v conspiracy. | '------------------------------^-------^------------------^--------------------'
PATCH 2/2 Enable paranoid path checking
In reply to Thomas PetazzoniHi Thomas,
Le 17/12/2014 22:47, Thomas Petazzoni a ?crit :
Quoted 9 lines
Now that Buildroot has support for paranoid path checking (check for clearly invalid header and library directories used when cross-compiling), let's enable this in the autobuilders.
Without this, the paranoid path checking only emits warnings. With this patch enabled, whenever an invalid header or library path is used, the build will stop and fail, which will allow us to clearly notice such cases.
Quoted 8 lines
Signed-off-by Thomas Petazzoni <[email protected]>---
Do we want to apply this right now? We already have a lot of failures caused by the static/shared change. Do we want to have more?Signed-off-by Thomas Petazzoni <[email protected]>---
Several packages have already been fixed since this feature has been added. Enabling this check now will help to discover the remaining packages left to fix.
Best regards,
Romain
PATCH 2/2 Enable paranoid path checking
Dear Romain Naour,
Quoted 5 lines · Romain Naour
Several packages have already been fixed since this feature has been added. Enabling this check now will help to discover the remaining packages left to fix.
Acked-by Romain Naour <[email protected]>
The patch has been committed, however the new version of the script has not yet been deployed on the autobuilders.
Thomas
Signature · 3 lines
Thomas Petazzoni, CTO, Free Electrons Embedded Linux, Kernel and Android engineering http://free-electrons.com
PATCH 2/2 Enable paranoid path checking
Hi Thomas,
Le 26/12/2014 19:09, Thomas Petazzoni a ?crit :
Quoted 8 lines
Dear Romain Naour,
Quoted 4 lines · Romain Naour
Several packages have already been fixed since this feature has been added. Enabling this check now will help to discover the remaining packages left to fix.
Quoted 5 lines
Quoted 1 line
Acked-by Romain Naour <[email protected]>The patch has been committed, however the new version of the script has not yet been deployed on the autobuilders.
Thomas
Ok, thanks.
I thought it was not because I didn't see any failure.
Best regards,
Romain