On Mon, 2012-03-19 at 14:35 -0700, Luis R. Rodriguez wrote: > From: "Luis R. Rodriguez" <mcgrof@xxxxxxxxxxxxx> > > It has happened twice now where elaborate troubleshooting has > undergone on systems where CONFIG_CFG80211_INTERNAL_REGDB [0] > has been set but yet net/wireless/db.txt was not updated. > > Despite the documentation on this it seems system integrators could > use some more help with this, so throw out a kernel warning at boot time > when their database is empty, and if they enabled cfg80211 regulatory > debugging (CONFIG_CFG80211_REG_DEBUG), simply fail at build time. If > they didn't enable cfg80211 regulatory debugging then we'll just warn > at boot time. This does mean that the error-prone system integrator > won't likely realize the issue until they boot the machine but -- it > does not seem to make sense to enable a build bug unless someone > really is debugging. > > [0] http://wireless.kernel.org/en/developers/Regulatory/CRDA#CONFIG_CFG80211_INTERNAL_REGDB > > Cc: Stephen Rothwell <sfr@xxxxxxxxxxxxxxxx> > Cc: Youngsin Lee <youngsin@xxxxxxxxxxxx> > Cc: Raja Mani <rmani@xxxxxxxxxxxxxxxx> > Cc: Senthil Kumar Balasubramanian <senthilb@xxxxxxxxxxxxxxxx> > Cc: Vipin Mehta <vipimeht@xxxxxxxxxxxxxxxx> > Cc: yahuan@xxxxxxxxxxxxxxxx > Cc: jjan@xxxxxxxxxxxxxxxx > Cc: vthiagar@xxxxxxxxxxxxxxxx > Cc: henrykim@xxxxxxxxxxxx > Cc: jouni@xxxxxxxxxxxxxxxx > Cc: athiruve@xxxxxxxxxxxxxxxx > Cc: cjkim@xxxxxxxxxxxx > Cc: philipk@xxxxxxxxxxxxxxxx > Cc: sunnykim@xxxxxxxxxxxx > Cc: sskwak@xxxxxxxxxxxx > Cc: kkim@xxxxxxxxxxxx > Cc: mattbyun@xxxxxxxxxxxx > Cc: ryanlee@xxxxxxxxxxxx > Cc: simbap@xxxxxxxxxxxx > Cc: krislee@xxxxxxxxxxxx > Cc: conner@xxxxxxxxxxxx > Cc: hojinkim@xxxxxxxxxxxx > Cc: honglee@xxxxxxxxxxxx > Cc: johnwkim@xxxxxxxxxxxx > Signed-off-by: Luis R. Rodriguez <mcgrof@xxxxxxxxxxxxx> > --- > > Stephen, John, not sure what is best to address this since I know we shouldn't > be failing to build kernels easily. This will only fail to build if both > CONFIG_CFG80211_REG_DEBUG and CONFIG_CFG80211_INTERNAL_REGDB are enabled. > We could get rid of the build bug check completely but this just means > system integrators will likley fail to fix the issue early. Experience > shows this is a common issue so far and a lot of time is going into debugging > this. I'd like to avoid this in the future. Let me know what you guys think. > > net/wireless/reg.c | 12 ++++++++++++ > 1 file changed, 12 insertions(+) > > diff --git a/net/wireless/reg.c b/net/wireless/reg.c > index e9a0ac8..85f51b3 100644 > --- a/net/wireless/reg.c > +++ b/net/wireless/reg.c > @@ -388,7 +388,18 @@ static void reg_regdb_query(const char *alpha2) > > schedule_work(®_regdb_work); > } > + > +/* Feel free to add any other sanity checks here */ > +static void reg_regdb_size_check(void) > +{ > +#ifdef CONFIG_CFG80211_REG_DEBUG > + BUILD_BUG_ON(!reg_regdb_size); > +#else > + WARN_ONCE(!reg_regdb_size, "db.txt is empty, you should update it..."); > +#endif That ifdef seems a bit pointless? If anything I would have expected it the other way around since the BUILD_BUG_ON compiles to nothing? johannes -- To unsubscribe from this list: send the line "unsubscribe linux-wireless" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
