Date: Sun, 30 May 2010 21:18:56 +0200 From: Marko Zec <zec@freebsd.org> To: Garrett Cooper <gcooper@freebsd.org> Cc: Perforce Change Reviews <perforce@freebsd.org> Subject: Re: PERFORCE change 178978 for review Message-ID: <201005302118.57240.zec@freebsd.org> In-Reply-To: <AANLkTilpZTmKr3tTrfafgt5CtskJB3LMEM_XoOsJdKq2@mail.gmail.com> References: <201005301444.o4UEidHb057496@repoman.freebsd.org> <AANLkTilpZTmKr3tTrfafgt5CtskJB3LMEM_XoOsJdKq2@mail.gmail.com>
index | next in thread | previous in thread | raw e-mail
On Sunday 30 May 2010 20:59:33 Garrett Cooper wrote: > On Sun, May 30, 2010 at 7:44 AM, Marko Zec <zec@freebsd.org> wrote: > > http://p4web.freebsd.org/@@178978?ac=10 > > > > Change 178978 by zec@zec_tpx32 on 2010/05/30 14:43:53 > > > > Attempt to reduce the diff against head a bit further and > > improve style, though Ivor hasn't left much here to be done... > > > > Affected files ... > > > > .. //depot/projects/vimage/src/sys/net/vnet.c#12 edit > > .. //depot/projects/vimage/src/sys/netinet/ip_mroute.c#42 edit > > > > Differences ... > > > > ==== //depot/projects/vimage/src/sys/net/vnet.c#12 (text+ko) ==== > > > > @@ -169,7 +169,7 @@ > > * we want the virtualized global variable space to be page-sized, we may > > * have more space than that in practice. > > */ > > -#define VNET_MODMIN (8192<<2) > > +#define VNET_MODMIN 32768 > > #define VNET_SIZE roundup2(VNET_BYTES, PAGE_SIZE) > > #define VNET_MODSIZE (VNET_SIZE - (VNET_BYTES - VNET_MODMIN)) > > > > > > ==== //depot/projects/vimage/src/sys/netinet/ip_mroute.c#42 (text+ko) > > ==== > > > > @@ -159,9 +159,9 @@ > > ((g).s_addr >> 20) ^ ((g).s_addr >> 10) ^ (g).s_addr) & > > V_mfchash) #define MFCHASHSIZE 256 > > > > -static VNET_DEFINE(u_char*, nexpire); > > +static u_long mfchashsize; /* Hash size */ > > +static VNET_DEFINE(u_char *, nexpire); /* 0..mfchashsize-1 */ > > #define V_nexpire VNET(nexpire) > > -static u_long mfchashsize; /* Hash size */ > > static VNET_DEFINE(LIST_HEAD(mfchashhdr, mfc)*, mfchashtbl); > > #define V_mfchashtbl VNET(mfchashtbl) > > > > @@ -218,7 +218,7 @@ > > */ > > static VNET_DEFINE(struct bw_upcall, bw_upcalls[BW_UPCALLS_MAX]); > > #define V_bw_upcalls VNET(bw_upcalls) > > -static VNET_DEFINE(u_int, bw_upcalls_n); /* # of pending upcalls */ > > +static VNET_DEFINE(u_int, bw_upcalls_n) = 0; /* # of pending upcalls */ > > #define V_bw_upcalls_n VNET(bw_upcalls_n) > > static VNET_DEFINE(struct callout, bw_upcalls_ch); > > #define V_bw_upcalls_ch VNET(bw_upcalls_ch) > > @@ -367,9 +367,9 @@ > > MRT_MFC_FLAGS_BORDER_VIF | > > MRT_MFC_RP | > > MRT_MFC_BW_UPCALL); > > -static VNET_DEFINE(uint32_t, mrt_api_config); > > +static VNET_DEFINE(uint32_t, mrt_api_config) = 0; > > #define V_mrt_api_config VNET(mrt_api_config) > > -static VNET_DEFINE(int, pim_assert_enabled); > > +static VNET_DEFINE(int, pim_assert_enabled) = 0; > > #define V_pim_assert_enabled VNET(pim_assert_enabled) > > static struct timeval pim_assert_interval = { 3, 0 }; /* Rate limit */ > > > > @@ -663,7 +663,7 @@ > > return EOPNOTSUPP; > > > > if (version != 1) > > - return (ENOPROTOOPT); > > + return ENOPROTOOPT; > > That's actually a style regression (along with any other return calls > without parentheses). Yup - technically you're right, but this particular change was intended to minimize the scope of the diff against head - if you'd throw a look at ip_mroute.c in head you'd spot a ton of other style violations, including this one. My intention is to do a functional (V_irtualization) change in one commit, and then we can do a no-op style sweep on ip_mroute.c later. Markohelp
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201005302118.57240.zec>