Skip to contentSkip to conversation

mt76: mt7615: Fix build with older compilers

PATCH In review Pablo Greco 3 participants 6 messages Dec 1, 2019 – Dec 2, 2019 1 message not in the archive

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

PATCH 1/1 mt76: mt7615: Fix build with older compilers

Some compilers (tested with 4.8.5 from CentOS 7) fail properly process FIELD_GET inside an inline function, which ends up in a BUILD_BUG_ON. Convert inline function to a macro.

Fixes commit bf92e7685100 ("mt76: mt7615: add support for per-chain signal strength reporting")
Reported in https://lkml.org/lkml/2019/9/21/146

---
 drivers/net/wireless/mediatek/mt76/mt7615/mac.c | 5 +----
 1 file changed, 1 insertion(+), 4 deletions(-)
drivers/net/wireless/mediatek/mt76/mt7615/mac.c +1 -4
index c77adc5d2552..77e395ca2c6a 100644
@@ -13,10 +13,7 @@
 #include "../dma.h"
 #include "mac.h"
 
-static inline s8 to_rssi(u32 field, u32 rxv)
-{
-	return (FIELD_GET(field, rxv) - 220) / 2;
-}
+#define to_rssi(field, rxv)		((FIELD_GET(field, rxv) - 220) / 2)
 
 static struct mt76_wcid *mt7615_rx_get_wcid(struct mt7615_dev *dev,
 					    u8 idx, bool unicast)
Signature · 7 lines
2.18.1


_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
#1 lore raw

Hello!

Quoted 1 line · Pablo Greco

Some compilers (tested with 4.8.5 from CentOS 7) fail properly process

Fail to?

Quoted 5 lines

FIELD_GET inside an inline function, which ends up in a BUILD_BUG_ON. Convert inline function to a macro.

Fixes commit bf92e7685100 ("mt76: mt7615: add support for per-chain signal strength reporting")

Should be:

Fixes bf92e7685100 ("mt76: mt7615: add support for per-chain signal strength reporting")

Do not ever break up the Fixes: line and don't insert empty lines between
it and other tags.

Quoted 4 lines

Reported in https://lkml.org/lkml/2019/9/21/146

[...]

MBR, Sergei

#2 lore raw
Quoted 7 lines · Sergei Shtylyov

Hello!

Quoted 1 line · Pablo Greco

Some compilers (tested with 4.8.5 from CentOS 7) fail properly process

   Fail to?

Right

Quoted 4 lines
Quoted 4 lines

FIELD_GET inside an inline function, which ends up in a BUILD_BUG_ON. Convert inline function to a macro.

Quoted 8 lines
Quoted 2 lines

Fixes commit bf92e7685100 ("mt76: mt7615: add support for per-chain signal strength reporting")

   Should be:

Fixes bf92e7685100 ("mt76: mt7615: add support for per-chain signal strength reporting")

   Do not ever break up the Fixes: line and don't insert empty lines between it and other tags.

Ack, I'll fix those for v2

Quoted 8 lines
Quoted 5 lines

Reported in https://lkml.org/lkml/2019/9/21/146

[...]

MBR, Sergei

Thanks, Pablo

#3 lore raw
In reply to Pablo Greco
Quoted 19 lines · Pablo Greco

Some compilers (tested with 4.8.5 from CentOS 7) fail properly process FIELD_GET inside an inline function, which ends up in a BUILD_BUG_ON. Convert inline function to a macro.

Fixes commit bf92e7685100 ("mt76: mt7615: add support for per-chain signal strength reporting")
Reported in https://lkml.org/lkml/2019/9/21/146

---
 drivers/net/wireless/mediatek/mt76/mt7615/mac.c | 5 +----
 1 file changed, 1 insertion(+), 4 deletions(-)
drivers/net/wireless/mediatek/mt76/mt7615/mac.c +0 -0
index c77adc5d2552..77e395ca2c6a 100644

drivers/net/wireless/mediatek/mt76/mt7615/mac.c line 16

Quoted 8 lines
drivers/net/wireless/mediatek/mt76/mt7615/mac.c +1 -4
@@ -13,10 +13,7 @@
 #include "../dma.h"
 #include "mac.h"
 
-static inline s8 to_rssi(u32 field, u32 rxv)
-{
-	return (FIELD_GET(field, rxv) - 220) / 2;
-}
+#define to_rssi(field, rxv)		((FIELD_GET(field, rxv) - 220) / 2)

What about u32_get_bits() instead of FIELD_GET(), would that work? I guess chances for that is slim, but it's always a shame to convert a function to a macro so we should try other methods first.

Or even better if we could fix FIELD_GET() to work with older compilers.

Signature · 1 line
https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches
#4 lore raw
In reply to Pablo Greco
Quoted 19 lines · Pablo Greco

Some compilers (tested with 4.8.5 from CentOS 7) fail properly process FIELD_GET inside an inline function, which ends up in a BUILD_BUG_ON. Convert inline function to a macro.

Fixes commit bf92e7685100 ("mt76: mt7615: add support for per-chain signal strength reporting")
Reported in https://lkml.org/lkml/2019/9/21/146

---
 drivers/net/wireless/mediatek/mt76/mt7615/mac.c | 5 +----
 1 file changed, 1 insertion(+), 4 deletions(-)
drivers/net/wireless/mediatek/mt76/mt7615/mac.c +0 -0
index c77adc5d2552..77e395ca2c6a 100644

drivers/net/wireless/mediatek/mt76/mt7615/mac.c line 16

Quoted 8 lines
drivers/net/wireless/mediatek/mt76/mt7615/mac.c +1 -4
@@ -13,10 +13,7 @@
 #include "../dma.h"
 #include "mac.h"
 
-static inline s8 to_rssi(u32 field, u32 rxv)
-{
-	return (FIELD_GET(field, rxv) - 220) / 2;
-}
+#define to_rssi(field, rxv)		((FIELD_GET(field, rxv) - 220) / 2)

What about u32_get_bits() instead of FIELD_GET(), would that work? I guess chances for that is slim, but it's always a shame to convert a function to a macro so we should try other methods first.

Or even better if we could fix FIELD_GET() to work with older compilers.

Signature · 6 lines
https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches

_______________________________________________
Linux-mediatek mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-mediatek

A message that is not in the archive (sent off-list or not yet received): 0101016ec5ed7d91-eac61501-1e4a-42f1-881d-cc2c02eb8372-000000@us-west-2.amazonses.com

#6 lore raw
Quoted 24 lines · Kalle Valo
Quoted 22 lines · Pablo Greco

Some compilers (tested with 4.8.5 from CentOS 7) fail properly process FIELD_GET inside an inline function, which ends up in a BUILD_BUG_ON. Convert inline function to a macro.

Fixes commit bf92e7685100 ("mt76: mt7615: add support for per-chain signal strength reporting")
Reported in https://lkml.org/lkml/2019/9/21/146

---
  drivers/net/wireless/mediatek/mt76/mt7615/mac.c | 5 +----
  1 file changed, 1 insertion(+), 4 deletions(-)
drivers/net/wireless/mediatek/mt76/mt7615/mac.c +0 -0
index c77adc5d2552..77e395ca2c6a 100644
@@ -13,10 +13,7 @@
  #include "../dma.h"
  #include "mac.h"
 
Quoted 8 lines

drivers/net/wireless/mediatek/mt76/mt7615/mac.c line 16

Quoted 5 lines
drivers/net/wireless/mediatek/mt76/mt7615/mac.c +1 -4
-static inline s8 to_rssi(u32 field, u32 rxv)
-{
-	return (FIELD_GET(field, rxv) - 220) / 2;
-}
+#define to_rssi(field, rxv)		((FIELD_GET(field, rxv) - 220) / 2)

What about u32_get_bits() instead of FIELD_GET(), would that work? I guess chances for that is slim, but it's always a shame to convert a function to a macro so we should try other methods first.

Anything that doesn't check field at build time should work, but between losing a check, or turning an inline into a macro, I'd rather use the macro.

Quoted 2 lines

Or even better if we could fix FIELD_GET() to work with older compilers.

The problem is not FIELD_GET itself, is that the compiler is trying to use "field" as a variable, instead as the macro expansion of GENMASK, as if the function wasn't inline. In the linked page you can see this message

BUILD_BUG_ON failed: (((field) + (1ULL << (__builtin_ffsll(field) - 1))) & (((field) + (1ULL << (__builtin_ffsll(field) - 1))) - 1)) != 0 _compiletime_assert(condition, msg, __compiletime_assert_, __LINE__)

which is is not right, because "field" should never be used for that check.

Pablo.