Date: Tue, 17 Dec 2019 13:28:20 -0500 From: Pedro Giffuni <pfg@FreeBSD.org> To: Cy Schubert <Cy.Schubert@cschubert.com> Cc: src-committers@freebsd.org, svn-src-all@freebsd.org, svn-src-head@freebsd.org Subject: Re: svn commit: r355828 - head/sys/sys Message-ID: <f6997e95-2b35-1975-2008-5b374195458b@FreeBSD.org> In-Reply-To: <201912170442.xBH4gUOg063777@slippy.cwsent.com> References: <201912162355.xBGNtUq6078840@repo.freebsd.org> <201912170442.xBH4gUOg063777@slippy.cwsent.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On 16/12/2019 23:42, Cy Schubert wrote: > In message <201912162355.xBGNtUq6078840@repo.freebsd.org>, "Pedro F. > Giffuni" w > rites: >> Author: pfg >> Date: Mon Dec 16 23:55:30 2019 >> New Revision: 355828 >> URL: https://svnweb.freebsd.org/changeset/base/355828 >> >> Log: >> Double the size of ARG_MAX on LP64 platforms. >> >> As modern software keeps growing in size, we get requests to update the >> value of ARG_MAX in order to link the resulting object files. Other OSs >> have much higher values but Increasiong ARG_MAX has a multiplied effect on >> KVA, so just bumping this value is dangerous in some archs like ARM32 that >> can exhaust KVA rather easily. >> >> While it would be better to have a unique value for all archs, other OSs >> (Illumos in partidular) can have different ARG_MAX limits depending on the >> platform, For now we want to be really conservative so we are avoidng >> the change on ILP32 and in the alternative case we only double it since tha >> t >> seems to work well enough for recent Code Aster. >> >> I was planning to bump the _FreeBSD_version but it was bumped recently >> (r355798) so we can reuse the 1300068 value for this change. > This doesn't seem right. Each bump should be for a distinct change and > documented as such. TBH, it is just not worth it: this change will currently benefit only one port (french/aster) and the update won't be committed until after the MFC is done. It is rather more painful that the change only fixes the case for some platforms. Pedro. > Also, it's not like we're saving any build time anyway. sys/syslimits.h > will cause a substantial portion to be rebuilt anyway. > >> >> PR: 241710 >> MFC after: 5 days >> >> Modified: >> head/sys/sys/syslimits.h >> >> Modified: head/sys/sys/syslimits.h >> ============================================================================= >> = >> --- head/sys/sys/syslimits.h Mon Dec 16 23:08:09 2019 (r355827) >> +++ head/sys/sys/syslimits.h Mon Dec 16 23:55:30 2019 (r355828) >> @@ -48,7 +48,11 @@ >> * Do not add any new variables here. (See the comment at the end of >> * the file for why.) >> */ >> -#define ARG_MAX 262144 /* max bytes for an exec functi >> on */ >> +#ifndef __ILP32__ >> +#define ARG_MAX (2 * 256 * 1024) /* max bytes for an exec functi >> on */ >> +#else >> +#define ARG_MAX (256 * 1024) /* max bytes for KVA-starved ar >> chs */ >> +#endif >> #ifndef CHILD_MAX >> #define CHILD_MAX 40 /* max simultaneous processes * >> / >> #endif >> >
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?f6997e95-2b35-1975-2008-5b374195458b>