{"thread":{"id":"9396","subject":"Problem with bisect","startedAt":"2007-08-05T16:02:21Z","lastAt":"2007-08-07T05:26:52Z","messageCount":9,"participants":["Larry Finger","Thomas Glanzmann","Sean","Christian Couder"],"isPatch":false,"patchVersion":null,"patchTotal":null},"messages":[{"id":"49907","messageId":"46B5F48D.7020907@lwfinger.net","threadId":"9396","inReplyTo":null,"subject":"Problem with bisect","fromName":"Larry Finger","fromEmail":"larry.finger@lwfinger.net","sentAt":"2007-08-05T16:02:21Z","receivedAt":"2007-08-05T16:02:21Z","isPatch":false,"sender":{"key":"larry.finger@lwfinger.net","avatar":null},"body":"I'm helping someone find what looks like a regression in bcm43xx-mac80211 between v2.6.22 and \nv2.6.23-rc1. This driver is not in the mainstream kernel, but is found in John Linville's \nwireless-dev git tree. When we do the first bisection between the current state and v2.6.22, we \nobtain a kernel whose Makefile says it is v2.6.22; however, it's code is based on a state before \nbcm43xx-mac80211 was introduced into this tree. My memory isn't what it used to be, but I think this \ncode was put into this tree during 2.6.19 or .20. When I used visualize to see the tree, the bottom \nis all the way to v2.6.16, which I think is the origin of the git process.\n\nIs this a git bug, or is it some flaw in this particular tree? We have worked around the problem by \narbitrarily calling each bisection that does not have the bcm43xx-mac80211 code as \"good\". It has \nbeen a source of confusion for the guy I'm helping as it is his first bisection. Unfortunately, the \nbug doesn't show on my machine.\n\nThanks,\n\nLarry\n"},{"id":"49929","messageId":"46B623D6.7070809@lwfinger.net","threadId":"9396","inReplyTo":"20070805145234.554bf671.seanlkml@sympatico.ca","subject":"Re: Problem with bisect","fromName":"Larry Finger","fromEmail":"larry.finger@lwfinger.net","sentAt":"2007-08-05T19:24:06Z","receivedAt":"2007-08-05T19:24:06Z","isPatch":false,"sender":{"key":"larry.finger@lwfinger.net","avatar":null},"body":"Sean wrote:\n> On Sun, 05 Aug 2007 11:02:21 -0500\n> Larry Finger wrote:\n> \n>> I'm helping someone find what looks like a regression in bcm43xx-mac80211 between v2.6.22 and \n>> v2.6.23-rc1. This driver is not in the mainstream kernel, but is found in John Linville's \n>> wireless-dev git tree. When we do the first bisection between the current state and v2.6.22, we \n>> obtain a kernel whose Makefile says it is v2.6.22; however, it's code is based on a state before \n>> bcm43xx-mac80211 was introduced into this tree. My memory isn't what it used to be, but I think this \n>> code was put into this tree during 2.6.19 or .20. When I used visualize to see the tree, the bottom \n>> is all the way to v2.6.16, which I think is the origin of the git process.\n>>\n>> Is this a git bug, or is it some flaw in this particular tree? We have worked around the problem by \n>> arbitrarily calling each bisection that does not have the bcm43xx-mac80211 code as \"good\". It has \n>> been a source of confusion for the guy I'm helping as it is his first bisection. Unfortunately, the \n>> bug doesn't show on my machine.\n>>\n\nThe git repo is git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless-dev.git.\n\nThe commands were:\n\ngit bisect start\ngit bisect bad\ngit bisect good v2.6.22\n\nI'm using git version 1.4.4.2.g04509\n\nLarry\n"},{"id":"49933","messageId":"20070805194454.GI25680@cip.informatik.uni-erlangen.de","threadId":"9396","inReplyTo":"46B623D6.7070809@lwfinger.net","subject":"Re: Problem with bisect","fromName":"Thomas Glanzmann","fromEmail":"thomas@glanzmann.de","sentAt":"2007-08-05T19:44:54Z","receivedAt":"2007-08-05T19:44:54Z","isPatch":false,"sender":{"key":"thomas@glanzmann.de","avatar":null},"body":"Hello,\n\n> I'm using git version 1.4.4.2.g04509\n\n1.4 has a issue with bisect at least the Debian 1.4 versio. Update to\n1.5 and try again.\n\n Thomas\n"},{"id":"49941","messageId":"46B633FC.5080004@lwfinger.net","threadId":"9396","inReplyTo":"20070805194454.GI25680@cip.informatik.uni-erlangen.de","subject":"Re: Problem with bisect","fromName":"Larry Finger","fromEmail":"larry.finger@lwfinger.net","sentAt":"2007-08-05T20:33:00Z","receivedAt":"2007-08-05T20:33:00Z","isPatch":false,"sender":{"key":"larry.finger@lwfinger.net","avatar":null},"body":"Thomas Glanzmann wrote:\n> Hello,\n> \n>> I'm using git version 1.4.4.2.g04509\n> \n> 1.4 has a issue with bisect at least the Debian 1.4 versio. Update to\n> 1.5 and try again.\n\nI'm now using git version 1.5.3.rc4 and still have the same situation.\n\nLarry\n"},{"id":"50034","messageId":"20070806141233.3d8a43cc.seanlkml@sympatico.ca","threadId":"9396","inReplyTo":"46B623D6.7070809@lwfinger.net","subject":"Re: Problem with bisect","fromName":"Sean","fromEmail":"seanlkml@sympatico.ca","sentAt":"2007-08-06T18:12:33Z","receivedAt":"2007-08-06T18:12:33Z","isPatch":false,"sender":{"key":"seanlkml@sympatico.ca","avatar":"https://gravatar.com/avatar/f92923f54fc08c401fc59b71829d4b89e9b8087fbba45ff87c82e6a83aee02ae?d=mp&s=160"},"body":"On Sun, 05 Aug 2007 14:24:06 -0500\nLarry Finger wrote:\n\n> Sean wrote:\n> > On Sun, 05 Aug 2007 11:02:21 -0500\n> > Larry Finger wrote:\n> > \n> >> I'm helping someone find what looks like a regression in bcm43xx-mac80211 between v2.6.22 and \n> >> v2.6.23-rc1. This driver is not in the mainstream kernel, but is found in John Linville's \n> >> wireless-dev git tree. When we do the first bisection between the current state and v2.6.22, we \n> >> obtain a kernel whose Makefile says it is v2.6.22; however, it's code is based on a state before \n> >> bcm43xx-mac80211 was introduced into this tree. My memory isn't what it used to be, but I think this \n> >> code was put into this tree during 2.6.19 or .20. When I used visualize to see the tree, the bottom \n> >> is all the way to v2.6.16, which I think is the origin of the git process.\n> >>\n> >> Is this a git bug, or is it some flaw in this particular tree? We have worked around the problem by \n> >> arbitrarily calling each bisection that does not have the bcm43xx-mac80211 code as \"good\". It has \n> >> been a source of confusion for the guy I'm helping as it is his first bisection. Unfortunately, the \n> >> bug doesn't show on my machine.\n> >>\n> The git repo is git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless-dev.git.\n> \n> The commands were:\n> \n> git bisect start\n> git bisect bad\n> git bisect good v2.6.22\n> \n> I'm using git version 1.4.4.2.g04509\n\nThe directory \"drivers/net/wireless/bcm43xx-mac80211\" is only introduced in\ncommit v2.6.23-rc1-1621-gd05daff. It didn't exist in v2.6.22.\n\nYou can see this with the command:\n\n $ git log -- drivers/net/wireless/bcm43xx-mac80211\n\nWhere the last listed commit is d05daff. So of course there will be many\nbisection points back to v2.6.22 where that directory just doesn't exist.\nA bit of digging with Git shows this history for most of the files in\nthat directory:\n\n renamed in v2.6.23-rc1-1621 as bcm43xx-mac80211\n renamed in v2.6.21-rc1-809 as mac80211\n renamed in v2.6.17-rc2-357 as d80211/bcm43xx\n Imported in v2.6.16-1725 as bcm43xx-d80211\n\nHTH,\nSean\n"},{"id":"50036","messageId":"46B76A7F.3000103@lwfinger.net","threadId":"9396","inReplyTo":"20070806141233.3d8a43cc.seanlkml@sympatico.ca","subject":"Re: Problem with bisect","fromName":"Larry Finger","fromEmail":"larry.finger@lwfinger.net","sentAt":"2007-08-06T18:37:51Z","receivedAt":"2007-08-06T18:37:51Z","isPatch":false,"sender":{"key":"larry.finger@lwfinger.net","avatar":null},"body":"Sean wrote:\n> \n> The directory \"drivers/net/wireless/bcm43xx-mac80211\" is only introduced in\n> commit v2.6.23-rc1-1621-gd05daff. It didn't exist in v2.6.22.\n> \n> You can see this with the command:\n> \n> $ git log -- drivers/net/wireless/bcm43xx-mac80211\n> \n> Where the last listed commit is d05daff. So of course there will be many\n> bisection points back to v2.6.22 where that directory just doesn't exist.\n> A bit of digging with Git shows this history for most of the files in\n> that directory:\n> \n> renamed in v2.6.23-rc1-1621 as bcm43xx-mac80211\n> renamed in v2.6.21-rc1-809 as mac80211\n> renamed in v2.6.17-rc2-357 as d80211/bcm43xx\n> Imported in v2.6.16-1725 as bcm43xx-d80211\n\nThanks for the response. Obviously the difficulties were due to the structure of the tree. This \nchunk of code has had a varied history.\n\nLarry\n"},{"id":"50067","messageId":"200708070350.50419.chriscool@tuxfamily.org","threadId":"9396","inReplyTo":"46B5F48D.7020907@lwfinger.net","subject":"Re: Problem with bisect","fromName":"Christian Couder","fromEmail":"chriscool@tuxfamily.org","sentAt":"2007-08-07T01:50:50Z","receivedAt":"2007-08-07T01:50:50Z","isPatch":false,"sender":{"key":"christian.couder@gmail.com","avatar":"https://avatars.githubusercontent.com/u/208954?v=4"},"body":"Le dimanche 5 août 2007 18:02, Larry Finger a écrit :\n> I'm helping someone find what looks like a regression in bcm43xx-mac80211\n> between v2.6.22 and v2.6.23-rc1. This driver is not in the mainstream\n> kernel, but is found in John Linville's wireless-dev git tree. When we do\n> the first bisection between the current state and v2.6.22, we obtain a\n> kernel whose Makefile says it is v2.6.22; however, it's code is based on\n> a state before bcm43xx-mac80211 was introduced into this tree. \n\nYou use \"git bisect good v2.6.22\", but this is not true because the \ntag \"v2.6.22\" is on the mainstream kernel branch and the driver is not \nthere.\n\nIf the v2.6.22 kernel that used to work came directly from John Linville's \nwireless-dev git tree, not from a patch, then you should find the exact \ncommit in John Linville's tree that worked and say \"git bisect good \".\n\nBut if the driver that worked with a mainstream v2.6.22 kernel had been \npatched, and now doesn't work when the same patch is applied to mainstream \nv2.6.23-rc1 kernel, then you can perhaps use:\n\ngit bisect start\ngit bisect bad v2.6.23-rc1\ngit bisect good v2.6.22\n\nand then:\n\n1) patch the kernel with the driver patch,\n2) test the patched kernel,\n3) remove the patch,\n4) say \"git bisect good\" or \"git bisect bad\"\n5) go to step 1) until the commit that broke the driver is found\n\nBest regards,\nChristian.\n"},{"id":"50082","messageId":"46B7DEB5.4060402@lwfinger.net","threadId":"9396","inReplyTo":"200708070350.50419.chriscool@tuxfamily.org","subject":"Re: Problem with bisect","fromName":"Larry Finger","fromEmail":"larry.finger@lwfinger.net","sentAt":"2007-08-07T02:53:41Z","receivedAt":"2007-08-07T02:53:41Z","isPatch":false,"sender":{"key":"larry.finger@lwfinger.net","avatar":null},"body":"Christian Couder wrote:\n> \n> You use \"git bisect good v2.6.22\", but this is not true because the \n> tag \"v2.6.22\" is on the mainstream kernel branch and the driver is not \n> there.\n> \n> If the v2.6.22 kernel that used to work came directly from John Linville's \n> wireless-dev git tree, not from a patch, then you should find the exact \n> commit in John Linville's tree that worked and say \"git bisect good commit>\".\n> \n> But if the driver that worked with a mainstream v2.6.22 kernel had been \n> patched, and now doesn't work when the same patch is applied to mainstream \n> v2.6.23-rc1 kernel, then you can perhaps use:\n> \n> git bisect start\n> git bisect bad v2.6.23-rc1\n> git bisect good v2.6.22\n> \n> and then:\n> \n> 1) patch the kernel with the driver patch,\n> 2) test the patched kernel,\n> 3) remove the patch,\n> 4) say \"git bisect good\" or \"git bisect bad\"\n> 5) go to step 1) until the commit that broke the driver is found\n\nHas the ability to use a commit hash to indicate a start point been in git for a long time? I think \nI remember trying it before when a version tag had not been downloaded and having a failure.\n\nLarry\n"},{"id":"50089","messageId":"200708070726.53428.chriscool@tuxfamily.org","threadId":"9396","inReplyTo":"46B7DEB5.4060402@lwfinger.net","subject":"Re: Problem with bisect","fromName":"Christian Couder","fromEmail":"chriscool@tuxfamily.org","sentAt":"2007-08-07T05:26:52Z","receivedAt":"2007-08-07T05:26:52Z","isPatch":false,"sender":{"key":"christian.couder@gmail.com","avatar":"https://avatars.githubusercontent.com/u/208954?v=4"},"body":"Le mardi 7 août 2007 04:53, Larry Finger a écrit :\n>\n> Has the ability to use a commit hash to indicate a start point been in\n> git for a long time?\n\nYes, it has always been there.\n\n> I think I remember trying it before when a version \n> tag had not been downloaded and having a failure.\n\nIt should have worked.\n\nBest regards,\nChristian.\n"}]}