Skip to contentSkip to conversation

modinfo crc32 broken after fd44a98ae2eb

DISCUSSION Samson Yeung 6 participants 9 messages Jan 27, 2016 – Jan 29, 2016

Subsystems (4)
lore
Share & tools
lore
#0 lore raw

Hi all,

Please reply directly as I am not on this list.

Commit "fd44a98ae2eb Fix race while loading modules" also broke 'modinfo crc32', specifically, for me.

Is this a bug, is it something strange going on with crc32, or is it intended behavior?

This happens on the official RedHat EL7 kernels, and also Ubuntu Wily's kernel.

Regards,
-Samson

#1 lore raw

Hi Samson,

Quoted 9 lines · Samson Yeung

Hi all,

Please reply directly as I am not on this list.

Commit "fd44a98ae2eb Fix race while loading modules" also broke 'modinfo crc32', specifically, for me.

Is this a bug, is it something strange going on with crc32, or is it intended behavior?

breaking is never intended ;-). What do you mean by "break"? Does it crash?

I can't see how this commit is related to modinfo. You will need to give more details. Check if you can get the output by passing the complete path to the module and make sure it's actually a module (and not builtin) in your kernel configuration

Lucas De Marchi

#2 lore raw
Quoted 7 lines · Lucas De Marchi

Hi Samson,

Quoted 4 lines · Samson Yeung

Hi all,

Please reply directly as I am not on this list.

Quoted 7 lines
Quoted 5 lines

Commit "fd44a98ae2eb Fix race while loading modules" also broke 'modinfo crc32', specifically, for me.

Is this a bug, is it something strange going on with crc32, or is it intended behavior?

breaking is never intended ;-). What do you mean by "break"? Does it crash?

'modinfo crc32' prints

modinfo: ERROR: Module crc32 not found.

and a return value of 1 using fd44a98ae2eb.

Without changing the contents of /lib/modules/uname -r/modules.* files, using fd44a98ae2eb^, 'modinfo crc32' gives me the expected info about crc32 instead:

filename:       /lib/modules/4.2.0-25-generic/kernel/crypto/crc32.ko
alias:          crypto-crc32
alias:          crc32
license:        GPL
description:    CRC32 calculations wrapper for lib/crc32
author:         Alexander Boyko <[email protected]>
srcversion:     B0DE005719AC9ED9A3EEAD7
depends:
intree:         Y
vermagic:       4.2.0-25-generic SMP mod_unload modversions
signer:         Build time autogenerated kernel key
sig_key:        A4:10:30:FB:DF:1D:C9:62:B4:BB:7D:16:44:C3:33:7E:C4:16:DB:86
sig_hashalgo:   sha512
Quoted 3 lines

I can't see how this commit is related to modinfo. You will need to give more details. Check if you can get the output by passing the complete path to the module and make sure it's actually a module (and

Giving the full path to the module still works as expected, so
'modinfo /lib/modules/4.2.0-25-generic/kernel/crypto/crc32.ko' displays module information and returns 0, using both fd44a98ae2eb^ and fd44a98ae2eb.

Quoted 1 line

not builtin) in your kernel configuration

I must admit that I'm slightly confused wrt crc32, since there is kernel/arch/x86/crypto/crc32-pclmul.ko as well as kernel/crypto/crc32.ko, and 'modinfo crc32-pclmul' claims that crc32 is an alias.

I This happens with a stock Ubuntu kernel (4.2.0-25-generic for Wily), however I first observed this behavior on Redhat/Oracle Linux 7.

Quoted 1 line

Lucas De Marchi

I have attached a slightly more complete log demonstrating the issue.

Regards,
-Samson

fragmede@biem:((tags/v20~8))~/tmp/kmod/git/kernel.org/kmod.git$ git checkout fd44a98ae2eb^ && ./autogen.sh >/dev/null && ./configure >/dev/null && make -j `nproc` >/dev/null && ./tools/modinfo crc32
Previous HEAD position was fd44a98ae2eb... Fix race while loading modules
HEAD is now at fae77516b405... build: we are kmod, not systemd
ar: `u' modifier ignored since `D' is the default (see `U')
filename:       /lib/modules/4.2.0-25-generic/kernel/crypto/crc32.ko
alias:          crypto-crc32
alias:          crc32
license:        GPL
description:    CRC32 calculations wrapper for lib/crc32
author:         Alexander Boyko <[email protected]>
srcversion:     B0DE005719AC9ED9A3EEAD7
depends:        
intree:         Y
vermagic:       4.2.0-25-generic SMP mod_unload modversions 
signer:         Build time autogenerated kernel key
sig_key:        A4:10:30:FB:DF:1D:C9:62:B4:BB:7D:16:44:C3:33:7E:C4:16:DB:86
sig_hashalgo:   sha512
fragmede@biem:((tags/v20~9))~/tmp/kmod/git/kernel.org/kmod.git$ git checkout fd44a98ae2eb && ./autogen.sh >/dev/null && ./configure >/dev/null && make -j `nproc` >/dev/null && ./tools/modinfo crc32
Previous HEAD position was fae77516b405... build: we are kmod, not systemd
HEAD is now at fd44a98ae2eb... Fix race while loading modules
ar: `u' modifier ignored since `D' is the default (see `U')
modinfo: ERROR: Module crc32 not found.
fragmede@biem:((tags/v20~8))~/tmp/kmod/git/kernel.org/kmod.git$ 
#3 lore raw

Hi!

CC'ing Rusty and mailing lists

Quoted 59 lines · Samson Yeung
Quoted 13 lines · Lucas De Marchi

Hi Samson,

Quoted 8 lines · Samson Yeung

Hi all,

Please reply directly as I am not on this list.

Commit "fd44a98ae2eb Fix race while loading modules" also broke 'modinfo crc32', specifically, for me.

Quoted 6 lines
Quoted 2 lines

Is this a bug, is it something strange going on with crc32, or is it intended behavior?

breaking is never intended ;-). What do you mean by "break"? Does it crash?

'modinfo crc32' prints

modinfo: ERROR: Module crc32 not found.

and a return value of 1 using fd44a98ae2eb.

Without changing the contents of /lib/modules/uname -r/modules.* files, using fd44a98ae2eb^, 'modinfo crc32' gives me the expected info about crc32 instead:

filename:       /lib/modules/4.2.0-25-generic/kernel/crypto/crc32.ko
alias:          crypto-crc32
alias:          crc32
license:        GPL
description:    CRC32 calculations wrapper for lib/crc32
author:         Alexander Boyko <[email protected]>
srcversion:     B0DE005719AC9ED9A3EEAD7
depends:
intree:         Y
vermagic:       4.2.0-25-generic SMP mod_unload modversions
signer:         Build time autogenerated kernel key
sig_key:        A4:10:30:FB:DF:1D:C9:62:B4:BB:7D:16:44:C3:33:7E:C4:16:DB:86
sig_hashalgo:   sha512
Quoted 3 lines

I can't see how this commit is related to modinfo. You will need to give more details. Check if you can get the output by passing the complete path to the module and make sure it's actually a module (and

Giving the full path to the module still works as expected, so 'modinfo /lib/modules/4.2.0-25-generic/kernel/crypto/crc32.ko' displays module information and returns 0, using both fd44a98ae2eb^ and fd44a98ae2eb.

Quoted 1 line

not builtin) in your kernel configuration

Quoted 6 lines

I must admit that I'm slightly confused wrt crc32, since there is kernel/arch/x86/crypto/crc32-pclmul.ko as well as kernel/crypto/crc32.ko, and 'modinfo crc32-pclmul' claims that crc32 is an alias.

I This happens with a stock Ubuntu kernel (4.2.0-25-generic for Wily), however I first observed this behavior on Redhat/Oracle Linux 7.

So... it looks like we have both kernel/lib/crc32.ko and kernel/crypto/crc32.ko - module names should be global. It doesn't clash in the kernel build because kernel/lib/crc32 is builtin most of the time.

With that commit in kmod there was a change on the builtin detection to fix the race. I'll take a look to fix this, but it should probably be fixed in the kernel
as well.

Rusty and ohers: it looks like both CONFIG_CRC32 and CONFIG_CRYPTO_CRC32 can be compiled as module, and they generate modules with the same name, crc32. Could that be fixed?

Lucas De Marchi

#4 lore raw

Duplicated module names

Quoted 3 lines · Lucas De Marchi

Hi!

CC'ing Rusty and mailing lists

Thanks.

Quoted 3 lines

Rusty and ohers: it looks like both CONFIG_CRC32 and CONFIG_CRYPTO_CRC32 can be compiled as module, and they generate modules with the same name, crc32. Could that be fixed?

Gah. Looks like it's been that way since at least 2014, too.

I think we could rename it to crypto_crc32, but I don't think it's the only one. Marco, I think depmod should probably FAIL if two modules have the same name, which would at least find such problems.

(BTW is there a nice way to figure out if a config var is a tristate? These are only problematic if both CONFIG_ are tristate.)

Here's a hacky attempt to look for problems:

rusty@rusty-Lemur:~/devel/kernel/linux (master)$ KCONFIGS=find * -name 'Kconfig*'; for m in find [b-z]* -name 'Makefile*'; do sed -n 's,obj-\$(CONFIG.*+= \([a-z0-9_-]\+\.o\)$,'$m' \1,p' <$m | sort -u; done | sort -k 2 | uniq -D -f 1 | while read m obj; do fgrep -w $obj $m /dev/null; done | while read LINE; do conf=echo $LINE | sed 's/.*\$(CONFIG_\([A-Z0-9_]*\).*/\1/'; if grep -C2 "^config $conf\$" $KCONFIGS | fgrep -q tristate; then echo $LINE; fi; done

Here are the results (mildly filtered by me):

drivers/gpu/drm/i2c/Makefile:obj-$(CONFIG_DRM_I2C_ADV7511) += adv7511.o drivers/media/i2c/Makefile:obj-$(CONFIG_VIDEO_ADV7511) += adv7511.o

drivers/media/platform/coda/Makefile:obj-$(CONFIG_VIDEO_CODA) += coda.o fs/coda/Makefile:obj-$(CONFIG_CODA_FS) += coda.o

drivers/gpu/drm/omapdrm/displays/Makefile:obj-$(CONFIG_DISPLAY_CONNECTOR_ANALOG_TV) += connector-analog-tv.o drivers/video/fbdev/omap2/omapfb/displays/Makefile:obj-$(CONFIG_FB_OMAP2_CONNECTOR_ANALOG_TV) += connector-analog-tv.o

drivers/gpu/drm/omapdrm/displays/Makefile:obj-$(CONFIG_DISPLAY_CONNECTOR_DVI) += connector-dvi.o drivers/video/fbdev/omap2/omapfb/displays/Makefile:obj-$(CONFIG_FB_OMAP2_CONNECTOR_DVI) += connector-dvi.o

drivers/gpu/drm/omapdrm/displays/Makefile:obj-$(CONFIG_DISPLAY_CONNECTOR_HDMI) += connector-hdmi.o drivers/video/fbdev/omap2/omapfb/displays/Makefile:obj-$(CONFIG_FB_OMAP2_CONNECTOR_HDMI) += connector-hdmi.o

crypto/Makefile:obj-$(CONFIG_CRYPTO_CRC32) += crc32.o lib/Makefile:obj-$(CONFIG_CRC32) += crc32.o

drivers/gpu/drm/omapdrm/displays/Makefile:obj-$(CONFIG_DISPLAY_ENCODER_OPA362) += encoder-opa362.o drivers/video/fbdev/omap2/omapfb/displays/Makefile:obj-$(CONFIG_FB_OMAP2_ENCODER_OPA362) += encoder-opa362.o

drivers/gpu/drm/omapdrm/displays/Makefile:obj-$(CONFIG_DISPLAY_ENCODER_TFP410) += encoder-tfp410.o drivers/video/fbdev/omap2/omapfb/displays/Makefile:obj-$(CONFIG_FB_OMAP2_ENCODER_TFP410) += encoder-tfp410.o

drivers/gpu/drm/omapdrm/displays/Makefile:obj-$(CONFIG_DISPLAY_ENCODER_TPD12S015) += encoder-tpd12s015.o drivers/video/fbdev/omap2/omapfb/displays/Makefile:obj-$(CONFIG_FB_OMAP2_ENCODER_TPD12S015) += encoder-tpd12s015.o

drivers/gpu/drm/omapdrm/displays/Makefile:obj-$(CONFIG_DISPLAY_PANEL_DPI) += panel-dpi.o drivers/video/fbdev/omap2/omapfb/displays/Makefile:obj-$(CONFIG_FB_OMAP2_PANEL_DPI) += panel-dpi.o

drivers/gpu/drm/omapdrm/displays/Makefile:obj-$(CONFIG_DISPLAY_PANEL_DSI_CM) += panel-dsi-cm.o drivers/video/fbdev/omap2/omapfb/displays/Makefile:obj-$(CONFIG_FB_OMAP2_PANEL_DSI_CM) += panel-dsi-cm.o

drivers/gpu/drm/omapdrm/displays/Makefile:obj-$(CONFIG_DISPLAY_PANEL_LGPHILIPS_LB035Q02) += panel-lgphilips-lb035q02.o drivers/video/fbdev/omap2/omapfb/displays/Makefile:obj-$(CONFIG_FB_OMAP2_PANEL_LGPHILIPS_LB035Q02) += panel-lgphilips-lb035q02.o

drivers/gpu/drm/omapdrm/displays/Makefile:obj-$(CONFIG_DISPLAY_PANEL_NEC_NL8048HL11) += panel-nec-nl8048hl11.o drivers/video/fbdev/omap2/omapfb/displays/Makefile:obj-$(CONFIG_FB_OMAP2_PANEL_NEC_NL8048HL11) += panel-nec-nl8048hl11.o

drivers/gpu/drm/omapdrm/displays/Makefile:obj-$(CONFIG_DISPLAY_PANEL_SHARP_LS037V7DW01) += panel-sharp-ls037v7dw01.o drivers/video/fbdev/omap2/omapfb/displays/Makefile:obj-$(CONFIG_FB_OMAP2_PANEL_SHARP_LS037V7DW01) += panel-sharp-ls037v7dw01.o

drivers/gpu/drm/omapdrm/displays/Makefile:obj-$(CONFIG_DISPLAY_PANEL_SONY_ACX565AKM) += panel-sony-acx565akm.o drivers/video/fbdev/omap2/omapfb/displays/Makefile:obj-$(CONFIG_FB_OMAP2_PANEL_SONY_ACX565AKM) += panel-sony-acx565akm.o

drivers/gpu/drm/omapdrm/displays/Makefile:obj-$(CONFIG_DISPLAY_PANEL_TPO_TD028TTEC1) += panel-tpo-td028ttec1.o drivers/video/fbdev/omap2/omapfb/displays/Makefile:obj-$(CONFIG_FB_OMAP2_PANEL_TPO_TD028TTEC1) += panel-tpo-td028ttec1.o

drivers/gpu/drm/omapdrm/displays/Makefile:obj-$(CONFIG_DISPLAY_PANEL_TPO_TD043MTEA1) += panel-tpo-td043mtea1.o drivers/video/fbdev/omap2/omapfb/displays/Makefile:obj-$(CONFIG_FB_OMAP2_PANEL_TPO_TD043MTEA1) += panel-tpo-td043mtea1.o

drivers/mtd/onenand/Makefile:obj-$(CONFIG_MTD_ONENAND_SAMSUNG) += samsung.o drivers/tty/serial/Makefile:obj-$(CONFIG_SERIAL_SAMSUNG) += samsung.o

sound/soc/codecs/Makefile:obj-$(CONFIG_SND_SOC_AC97_CODEC) += snd-soc-ac97.o sound/soc/samsung/Makefile:obj-$(CONFIG_SND_SAMSUNG_AC97) += snd-soc-ac97.o

drivers/hwmon/Makefile:obj-$(CONFIG_SENSORS_VEXPRESS) += vexpress.o drivers/regulator/Makefile:obj-$(CONFIG_REGULATOR_VEXPRESS) += vexpress.o

Can maintainers please tell me which of these are harmless?

Thanks!
Rusty.

#5 lore raw

Re: Duplicated module names

Quoted 80 lines · Rusty Russell
Quoted 3 lines · Lucas De Marchi

Hi!

CC'ing Rusty and mailing lists

Thanks.

Quoted 3 lines

Rusty and ohers: it looks like both CONFIG_CRC32 and CONFIG_CRYPTO_CRC32 can be compiled as module, and they generate modules with the same name, crc32. Could that be fixed?

Gah. Looks like it's been that way since at least 2014, too.

I think we could rename it to crypto_crc32, but I don't think it's the only one. Marco, I think depmod should probably FAIL if two modules have the same name, which would at least find such problems.

(BTW is there a nice way to figure out if a config var is a tristate? These are only problematic if both CONFIG_ are tristate.)

Here's a hacky attempt to look for problems:

rusty@rusty-Lemur:~/devel/kernel/linux (master)$ KCONFIGS=find * -name 'Kconfig*'; for m in find [b-z]* -name 'Makefile*'; do sed -n 's,obj-\$(CONFIG.*+= \([a-z0-9_-]\+\.o\)$,'$m' \1,p' <$m | sort -u; done | sort -k 2 | uniq -D -f 1 | while read m obj; do fgrep -w $obj $m /dev/null; done | while read LINE; do conf=echo $LINE | sed 's/.*\$(CONFIG_\([A-Z0-9_]*\).*/\1/'; if grep -C2 "^config $conf\$" $KCONFIGS | fgrep -q tristate; then echo $LINE; fi; done

Here are the results (mildly filtered by me):

drivers/gpu/drm/i2c/Makefile:obj-$(CONFIG_DRM_I2C_ADV7511) += adv7511.o drivers/media/i2c/Makefile:obj-$(CONFIG_VIDEO_ADV7511) += adv7511.o

drivers/media/platform/coda/Makefile:obj-$(CONFIG_VIDEO_CODA) += coda.o fs/coda/Makefile:obj-$(CONFIG_CODA_FS) += coda.o

drivers/gpu/drm/omapdrm/displays/Makefile:obj-$(CONFIG_DISPLAY_CONNECTOR_ANALOG_TV) += connector-analog-tv.o drivers/video/fbdev/omap2/omapfb/displays/Makefile:obj-$(CONFIG_FB_OMAP2_CONNECTOR_ANALOG_TV) += connector-analog-tv.o

drivers/gpu/drm/omapdrm/displays/Makefile:obj-$(CONFIG_DISPLAY_CONNECTOR_DVI) += connector-dvi.o drivers/video/fbdev/omap2/omapfb/displays/Makefile:obj-$(CONFIG_FB_OMAP2_CONNECTOR_DVI) += connector-dvi.o

drivers/gpu/drm/omapdrm/displays/Makefile:obj-$(CONFIG_DISPLAY_CONNECTOR_HDMI) += connector-hdmi.o drivers/video/fbdev/omap2/omapfb/displays/Makefile:obj-$(CONFIG_FB_OMAP2_CONNECTOR_HDMI) += connector-hdmi.o

crypto/Makefile:obj-$(CONFIG_CRYPTO_CRC32) += crc32.o lib/Makefile:obj-$(CONFIG_CRC32) += crc32.o

drivers/gpu/drm/omapdrm/displays/Makefile:obj-$(CONFIG_DISPLAY_ENCODER_OPA362) += encoder-opa362.o drivers/video/fbdev/omap2/omapfb/displays/Makefile:obj-$(CONFIG_FB_OMAP2_ENCODER_OPA362) += encoder-opa362.o

drivers/gpu/drm/omapdrm/displays/Makefile:obj-$(CONFIG_DISPLAY_ENCODER_TFP410) += encoder-tfp410.o drivers/video/fbdev/omap2/omapfb/displays/Makefile:obj-$(CONFIG_FB_OMAP2_ENCODER_TFP410) += encoder-tfp410.o

drivers/gpu/drm/omapdrm/displays/Makefile:obj-$(CONFIG_DISPLAY_ENCODER_TPD12S015) += encoder-tpd12s015.o drivers/video/fbdev/omap2/omapfb/displays/Makefile:obj-$(CONFIG_FB_OMAP2_ENCODER_TPD12S015) += encoder-tpd12s015.o

drivers/gpu/drm/omapdrm/displays/Makefile:obj-$(CONFIG_DISPLAY_PANEL_DPI) += panel-dpi.o drivers/video/fbdev/omap2/omapfb/displays/Makefile:obj-$(CONFIG_FB_OMAP2_PANEL_DPI) += panel-dpi.o

drivers/gpu/drm/omapdrm/displays/Makefile:obj-$(CONFIG_DISPLAY_PANEL_DSI_CM) += panel-dsi-cm.o drivers/video/fbdev/omap2/omapfb/displays/Makefile:obj-$(CONFIG_FB_OMAP2_PANEL_DSI_CM) += panel-dsi-cm.o

drivers/gpu/drm/omapdrm/displays/Makefile:obj-$(CONFIG_DISPLAY_PANEL_LGPHILIPS_LB035Q02) += panel-lgphilips-lb035q02.o drivers/video/fbdev/omap2/omapfb/displays/Makefile:obj-$(CONFIG_FB_OMAP2_PANEL_LGPHILIPS_LB035Q02) += panel-lgphilips-lb035q02.o

drivers/gpu/drm/omapdrm/displays/Makefile:obj-$(CONFIG_DISPLAY_PANEL_NEC_NL8048HL11) += panel-nec-nl8048hl11.o drivers/video/fbdev/omap2/omapfb/displays/Makefile:obj-$(CONFIG_FB_OMAP2_PANEL_NEC_NL8048HL11) += panel-nec-nl8048hl11.o

drivers/gpu/drm/omapdrm/displays/Makefile:obj-$(CONFIG_DISPLAY_PANEL_SHARP_LS037V7DW01) += panel-sharp-ls037v7dw01.o drivers/video/fbdev/omap2/omapfb/displays/Makefile:obj-$(CONFIG_FB_OMAP2_PANEL_SHARP_LS037V7DW01) += panel-sharp-ls037v7dw01.o

drivers/gpu/drm/omapdrm/displays/Makefile:obj-$(CONFIG_DISPLAY_PANEL_SONY_ACX565AKM) += panel-sony-acx565akm.o drivers/video/fbdev/omap2/omapfb/displays/Makefile:obj-$(CONFIG_FB_OMAP2_PANEL_SONY_ACX565AKM) += panel-sony-acx565akm.o

drivers/gpu/drm/omapdrm/displays/Makefile:obj-$(CONFIG_DISPLAY_PANEL_TPO_TD028TTEC1) += panel-tpo-td028ttec1.o drivers/video/fbdev/omap2/omapfb/displays/Makefile:obj-$(CONFIG_FB_OMAP2_PANEL_TPO_TD028TTEC1) += panel-tpo-td028ttec1.o

drivers/gpu/drm/omapdrm/displays/Makefile:obj-$(CONFIG_DISPLAY_PANEL_TPO_TD043MTEA1) += panel-tpo-td043mtea1.o drivers/video/fbdev/omap2/omapfb/displays/Makefile:obj-$(CONFIG_FB_OMAP2_PANEL_TPO_TD043MTEA1) += panel-tpo-td043mtea1.o

drivers/mtd/onenand/Makefile:obj-$(CONFIG_MTD_ONENAND_SAMSUNG) += samsung.o drivers/tty/serial/Makefile:obj-$(CONFIG_SERIAL_SAMSUNG) += samsung.o

Quoted 7 lines

sound/soc/codecs/Makefile:obj-$(CONFIG_SND_SOC_AC97_CODEC) += snd-soc-ac97.o sound/soc/samsung/Makefile:obj-$(CONFIG_SND_SAMSUNG_AC97) += snd-soc-ac97.o

drivers/hwmon/Makefile:obj-$(CONFIG_SENSORS_VEXPRESS) += vexpress.o drivers/regulator/Makefile:obj-$(CONFIG_REGULATOR_VEXPRESS) += vexpress.o

Can maintainers please tell me which of these are harmless?

The duplicates under drivers/gpu/drm/omapdrm/ and drivers/video/fbdev/omap2/ are (hopefully) harmless, as the DRM driver and the fbdev driver are mutually exclusive.

Tomi

maintainer
#6 lore raw

Re: Duplicated module names

In reply to Rusty Russell
Quoted 4 lines · Rusty Russell

I think we could rename it to crypto_crc32, but I don't think it's the only one. Marco, I think depmod should probably FAIL if two modules have the same name, which would at least find such problems.

This patch renames it to crc32_generic.

---8<---
Subject: crypto: crc32 - Rename generic implementation

The generic crc32 implementation is currently called crc32. This is a problem because it clashes with the lib implementation of crc32.

This patch renames the crypto crc32 to crc32_generic so that it is consistent with crc32c. An alias for the driver is also added.

crypto/Makefile +1 -1
index ffe18c9..059de1b 100644
@@ -100,7 +100,7 @@ obj-$(CONFIG_CRYPTO_POLY1305) += poly1305_generic.o
 obj-$(CONFIG_CRYPTO_DEFLATE) += deflate.o
 obj-$(CONFIG_CRYPTO_MICHAEL_MIC) += michael_mic.o
 obj-$(CONFIG_CRYPTO_CRC32C) += crc32c_generic.o
-obj-$(CONFIG_CRYPTO_CRC32) += crc32.o
+obj-$(CONFIG_CRYPTO_CRC32) += crc32_generic.o
 obj-$(CONFIG_CRYPTO_CRCT10DIF) += crct10dif_common.o crct10dif_generic.o
 obj-$(CONFIG_CRYPTO_AUTHENC) += authenc.o authencesn.o
 obj-$(CONFIG_CRYPTO_LZO) += lzo.o
crypto/crc32_generic.c renamed from crypto/crc32.c +2 -1
similarity index 98%
rename from crypto/crc32.c
rename to crypto/crc32_generic.c
index 187ded2..aa2a25f 100644
@@ -131,7 +131,7 @@ static struct shash_alg alg = {
 	.digestsize	= CHKSUM_DIGEST_SIZE,
 	.base		= {
 		.cra_name		= "crc32",
-		.cra_driver_name	= "crc32-table",
+		.cra_driver_name	= "crc32-generic",
 		.cra_priority		= 100,
 		.cra_blocksize		= CHKSUM_BLOCK_SIZE,
 		.cra_ctxsize		= sizeof(u32),
@@ -157,3 +157,4 @@ MODULE_AUTHOR("Alexander Boyko <[email protected]>");
 MODULE_DESCRIPTION("CRC32 calculations wrapper for lib/crc32");
 MODULE_LICENSE("GPL");
 MODULE_ALIAS_CRYPTO("crc32");
+MODULE_ALIAS_CRYPTO("crc32-generic");
Signature · 3 lines
Email: Herbert Xu <[email protected]>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
#7 lore raw

Re: Duplicated module names

In reply to Rusty Russell

On Fri, Jan 29, 2016 at 5:54 AM, Rusty Russell <[email protected]> wrot= e:

Quoted 11 lines
Quoted 3 lines · Lucas De Marchi

Hi!

CC'ing Rusty and mailing lists

Thanks.

Quoted 3 lines

Rusty and ohers: it looks like both CONFIG_CRC32 and CONFIG_CRYPTO_CRC32 can be compiled as module, and they generate modules with the same name, crc32. Could that be fixed?

Quoted 7 lines

Gah. Looks like it's been that way since at least 2014, too.

I think we could rename it to crypto_crc32, but I don't think it's the only one. Marco, I think depmod should probably FAIL if two modules have the same name, which would at least find such problems.

(BTW is there a nice way to figure out if a config var is a tristate? Th=

ese

Quoted 5 lines

are only problematic if both CONFIG_ are tristate.)

Here's a hacky attempt to look for problems:

rusty@rusty-Lemur:~/devel/kernel/linux (master)$ KCONFIGS=3D`find * -name=

'Kconfig*'; for m in find [b-z]* -name 'Makefile*'`; do sed -n 's,obj-\$=
(CONFIG.*+=3D \([a-z0-9_-]\+\.o\)$,'$m' \1,p' <$m | sort -u; done | sort -k=
2 | uniq -D -f 1 | while read m obj; do fgrep -w $obj $m /dev/null; done |= while read LINE; do conf=3D`echo $LINE | sed 's/.*\$(CONFIG_\([A-Z0-9_]*\)=
.*/\1/'`; if grep -C2 "^config $conf\$" $KCONFIGS | fgrep -q tristate; then=
echo $LINE; fi; done

Quoted 3 lines

Here are the results (mildly filtered by me):

[...]

Quoted 3 lines

drivers/hwmon/Makefile:obj-$(CONFIG_SENSORS_VEXPRESS) +=3D vexpress.o drivers/regulator/Makefile:obj-$(CONFIG_REGULATOR_VEXPRESS) +=3D vexpress=

.o

Quoted 1 line

Indeed causes issues if both are built as modules, only once succeeds. I will post these 2 patches separately soon.

-->8

Quoted 1 line

From 1eb0d653d90e3f5fe05c6f63a17976226309e1ec Mon Sep 17 00:00:00 2001

From: Sudeep Holla <[email protected]>
Date: Fri, 29 Jan 2016 15:40:24 +0000
Subject: [PATCH 1/2] hwmon: (vexpress) rename vexpress hwmon implementation

The vexpress hwmon implementation is currently just called vexpress. This is a problem because it clashes with another module with the same name in regulators.

This patch renames the vexpress hwmon implementation to vexpress-hwmon so that there will be no clash in the module namespace.

---
 drivers/hwmon/Makefile                         | 2 +-
 drivers/hwmon/{vexpress.c =3D> vexpress-hwmon.c} | 0
 2 files changed, 1 insertion(+), 1 deletion(-)
 rename drivers/hwmon/{vexpress.c =3D> vexpress-hwmon.c} (100%)
drivers/hwmon/Makefile +1 -1
index 30c94df31465..cfc09711810c 100644
@@ -149,7 +149,7 @@ obj-$(CONFIG_SENSORS_TMP103) +=3D tmp103.o
 obj-$(CONFIG_SENSORS_TMP401) +=3D tmp401.o
 obj-$(CONFIG_SENSORS_TMP421) +=3D tmp421.o
 obj-$(CONFIG_SENSORS_TWL4030_MADC)+=3D twl4030-madc-hwmon.o
-obj-$(CONFIG_SENSORS_VEXPRESS) +=3D vexpress.o
+obj-$(CONFIG_SENSORS_VEXPRESS) +=3D vexpress-hwmon.o
 obj-$(CONFIG_SENSORS_VIA_CPUTEMP)+=3D via-cputemp.o
 obj-$(CONFIG_SENSORS_VIA686A) +=3D via686a.o
 obj-$(CONFIG_SENSORS_VT1211) +=3D vt1211.o
drivers/hwmon/vexpress-hwmon.c renamed from drivers/hwmon/vexpress.c +0 -0
similarity index 100%
rename from drivers/hwmon/vexpress.c
rename to drivers/hwmon/vexpress-hwmon.c

--=20
1.9.1

Quoted 1 line

From 1520bd20d3863df87ca5d9461bf5a20d3f58e385 Mon Sep 17 00:00:00 2001

From: Sudeep Holla <[email protected]>
Date: Fri, 29 Jan 2016 15:44:50 +0000
Subject: [PATCH 2/2] regulator: vexpress: rename vexpress regulator
implementation

The vexpress regulator implementation is currently just called vexpress. This is a problem because it clashes with another module with the same name in hardware monitors.

This patch renames the vexpress regulator implementation to vexpress-regulator so that there will be no clash in the module namespace.

---
 drivers/regulator/Makefile                             | 2 +-
 drivers/regulator/{vexpress.c =3D> vexpress-regulator.c} | 0
 2 files changed, 1 insertion(+), 1 deletion(-)
 rename drivers/regulator/{vexpress.c =3D> vexpress-regulator.c} (100%)
drivers/regulator/Makefile +0 -0
index 980b1943fa81..755077a89a25 100644
@@ -98,7 +98,7 @@ obj-$(CONFIG_REGULATOR_TPS65910) +=3D tps65910-regulator.=

o
obj-$(CONFIG_REGULATOR_TPS65912) +=3D tps65912-regulator.o obj-$(CONFIG_REGULATOR_TPS80031) +=3D tps80031-regulator.o obj-$(CONFIG_REGULATOR_TWL4030) +=3D twl-regulator.o
-obj-$(CONFIG_REGULATOR_VEXPRESS) +=3D vexpress.o +obj-$(CONFIG_REGULATOR_VEXPRESS) +=3D vexpress-regulator.o
obj-$(CONFIG_REGULATOR_WM831X) +=3D wm831x-dcdc.o obj-$(CONFIG_REGULATOR_WM831X) +=3D wm831x-isink.o obj-$(CONFIG_REGULATOR_WM831X) +=3D wm831x-ldo.o

(quoted diff) +0 -0

b/drivers/regulator/vexpress-regulator.c
similarity index 100%
rename from drivers/regulator/vexpress.c
rename to drivers/regulator/vexpress-regulator.c
--=20
1.9.1

#8 lore raw

Re: Duplicated module names

In reply to Rusty Russell

Hi!

Quoted 11 lines · Rusty Russell
Quoted 3 lines · Lucas De Marchi

Hi!

CC'ing Rusty and mailing lists

Thanks.

Quoted 3 lines

Rusty and ohers: it looks like both CONFIG_CRC32 and CONFIG_CRYPTO_CRC32 can be compiled as module, and they generate modules with the same name, crc32. Could that be fixed?

Quoted 5 lines

Gah. Looks like it's been that way since at least 2014, too.

I think we could rename it to crypto_crc32, but I don't think it's the only one. Marco, I think depmod should probably FAIL if two modules have the same name, which would at least find such problems.

Yes, I'll do that on kmod, but we need to pay attention to 2 things:

- It's possible 2 modules with the same names if they come from
different locations (with different priorities)
- It doesn't fix the case in which one is a module and the other is
builtin. We could still have problem with such a scenario, like was pointed out in this bug report to kmod.

Lucas De Marchi